Show Notes:
1:00 Anki Sanhi
1:45 Sahni’s background and work
5:00 Eric Raymond’s The Cathedral and the Bazaar - essay on open source development
6:40 association with the arts
7:45 trademark enforcement work
10:00 infringement case between client working in fashion space sued by global brand
12:30 origin of the Raghav project
16:45 process to use Raghav
18:30 Van Gogh’s Starry Night
21:00 Raghav = Robust Artificially Intelligent Graphics and Arts Visualizer
21:45 Indian copyright application for “Suryast”
25:40 Section 2(d) of Indian Copyright Act, definition of author
34:30 Canadian copyright application for “Suryast”
35:35 Notice of Application filed in Canadian Federal Court by Samuelson-Glushko Canadian Internet Policy and Public Interest Clinic (CIPPIC) against Sahni over “Suryast”
37:10 US copyright application for “Suryast”
37:35 Sahni’s US Counsel, Alex Garens, Esq. with Day Pitney
38:00 USCO decision on “Suryast”
43:30 India’s definition of ‘author’ - no qualifications
46:50 Parliamentary Standing Committee recommended creation of new right for AI works per Sahni’s recommendation
47:05 February 2024 - Union Minister of State Commerce in India’s Parliament submitted that there is no need for creation of a new right and current legislation offers sufficient protection
48:45 ambiguity in Canada law on AI
49:40 USCO’s Request for Comments
52:00 Sahni’s position of need for amendment of US law to address AI
53:50 Optimus
54:50 Definition of justice
56:30 global harmonized principles on AI
57:30 injustice of humans unknowingly competing against AI due to historic suppression of AI use
59:00 Sahni’s intent to contribute to the change in law to address AI
1:00:00 consultation with Ryan Abbott, Esq.
1:00:50 Future projects
1:02:20 shaping issues where art and law intersect
1:03:30 cultural impact of the debate over human authorship, AI and art
1:04:30 comparison of photography to AI
Please share your comments and/or questions at stephanie@warfareofartandlaw.com
Music by Toulme.
To hear more episodes, please visit Warfare of Art and Law podcast's website.
To leave questions or comments about this or other episodes of the podcast and/or for information about joining the 2ND Saturday discussion on art, culture and justice, please message me at stephanie@warfareofartandlaw.com.
Thanks so much for listening!
© Stephanie Drawdy [2025]
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:02):
So I think we should embrace technology, as humans
have always had, withoutcarrying any apprehensions
regarding any negative thingsthat might impact culture or
society as a whole.
Speaker 2 (00:18):
Welcome to Warfare of Art and Law, the podcast that
focuses on how justice does ordoesn't play out when art and
law overlap.
Hi everyone, it's Stephanie,and that was intellectual
property and technology lawsattorney Ankit Sanhi.
In the following conversation,we discuss what drew him to IP
(00:40):
and tech law and the motivationbehind the Raghav Project and
tech law and the motivationbehind the Raghav project.
We also discuss the range ofresponses he's received from his
copyright applications to India, canada and the US for
registration of the Raghavoutput entitled Suriast, and we
(01:00):
close with considerations aboutthe definition of justice and
the cultural impact of emergingtech overlapping with art.
Ankit Sanhi, welcome to Warfareof Art and Law.
Thank you so much for beinghere.
Speaker 1 (01:19):
It's a pleasure, stephanie.
Thank you so much for giving methe opportunity to be on your
podcast, and I'm really lookingforward to our conversation
today.
Speaker 2 (01:28):
As am I, and perhaps we start with you giving a bit
about your background and yourwork.
Speaker 1 (01:36):
Sure.
So you know I've had a rathercliched background and career so
far as IP and technology lawsis concerned.
I'm a second generationtechnology and IP lawyer, so it
started very early, and almostin middle school is when I
(02:00):
nearly made up my mind that I'mgoing to go in the direction of
pursuing a law degree withspecialization in intellectual
property, innovation andtechnology.
I was always deeply passionateabout technology.
I was also my school's computersociety's president.
(02:23):
Back then this was theterminology that we would use
for anything which wastech-related.
So it was our school's computersociety, and I would basically
be very much engaged in allkinds of inter-school
competitions pertaining toprogramming and quizzing and so
(02:47):
on on various aspects oftechnology.
I also, incidentally, was oneof India's youngest to qualify
Microsoft certifiedprofessionals at the age of 14.
And so it kind of almost wasnatural for me to get an
(03:07):
opportunity to be where I am, asI'd always thought of being in
this kind of role well, that issuch a rich background.
Speaker 2 (03:20):
Like you are steeped in this work, and so I do wonder
, like you had opinions early onabout IP protection, can you
kind of walk us through that andhow they've evolved?
Speaker 1 (03:37):
Sure.
So by this do you mean how theconcept of IP protection may
have evolved?
Do you mean how?
Speaker 2 (03:44):
the concept of IP protection may have evolved.
Yeah well, just like youropinion.
Like some people are opensource of the opinion that open
source is the way to go, andothers are very steeped in
protecting and keeping it safefrom use.
So I wonder, like, has therebeen an arc in how you've seen
(04:07):
that or how is that for you?
Speaker 1 (04:11):
Yes, in fact, I'm so glad you brought up the question
.
You know, there's one thingthat I noticed in this journey,
and while we were growing up andyou know I went to law school,
and while we were growing up andI went to law school when we
start off as students, we'renaturally given various
perspectives, and the credit tothat goes to our teachers.
(04:35):
And what I've learned is, whenyou start out, the purpose of
really pursuing a law degree asyou may have felt yourself the
same way at some stage, Ibelieve is really to do justice
and your perspective's a bitmore leaning on the liberal side
(04:57):
than on the capitalist side,because that is the kind of
exposure, the academic exposure,that you get.
So, you know, one of the earlyworks that inspired me was Eric
Raymond's the Cathedral and theBazaar, and I was deeply
fascinated by the world of opensource.
(05:20):
Wikipedia had just become athing, wikipedia had just become
(05:49):
a thing, and for a model likethat to exist on global
contribution amongst a set ofpeople, individual parts that
they contributed, was thebiggest fascination.
But then, you know, the devilgets the best of us, as I like
to say.
And when you join theprofession, when you get the
opportunity to representcompanies, large corporations,
(06:09):
and defend their rights in courtor enforce their rights in
court of the spectrum all theway to pro-protection and
pro-right holders, becauseyou're really fighting for your
clients and representing themeach day in court or before the
(06:31):
IP office and so on.
Speaker 2 (06:38):
Yeah, absolutely, and I also.
Just as an aside, did you alsohave a background in the arts,
or like, what is your exposureand involvement with the arts?
Speaker 1 (06:54):
So I've always been passionate about art.
In general, I've had a thingfor theater and drama.
I used to be part of a theatergroup when I was in school, so
I've always had a lot ofadmiration for artists of our
era, artists of the times, theeras that have gone by, and in
(07:17):
that sense I remain associatedwith the arts.
But besides that, I haven'treally had any formal education
or training in the arts field assuch.
Speaker 2 (07:31):
Well, so then, I guess I would be curious to know
, like this, this law and techwork that you've been doing like
.
Are there certain cases thatexcuse me, certain cases that
stand out to you as beingemblematic of the work you do,
or certain ones you might wantto highlight?
Speaker 1 (07:57):
I've done more of cases on the trademark law side.
I'm primarily a trademarkenforcement person, if I could
put it that way.
Of course, we all have had ourfair share of contesting for
clients in design and copyrightcases as well.
I am fortunate to represent afew leading music artists and
(08:23):
some leading painters from India, and we represent their
interests both commercially aswell as in terms of protecting
their personality rights andother aspects.
But largely as a trademarklawyer, we've spent more time
defending large brands frombeing counterfeited or from bad
(08:49):
faith registrations or attemptsat passing off.
So I wouldn't say that myprofessional journey has really
had much association with theRaghav project or with Suryast
as such.
But you know there have alwaysbeen interesting matters.
I can probably spend a minuteor so talking about one of the
(09:15):
interesting mandates that we got, because it involves certain
aspects which are stillconfidential, to the extent that
that particular matter is stilllive.
I will refrain from using names, but I will try to fill in on
the interesting tidbits.
So we basically got a mandatefrom a large Indian
(09:38):
multinational corporation thatwas interested in, that was
interested in acquiring masterrights, licensee rights to a
large global brand, a retailbrand, and they came to us, the
business folks, saying that theytried their best to impress the
(10:00):
guys back in the United Statesfor giving them the exclusivity
for brand rights in India andthe subcontinent, but couldn't
succeed.
So their strategy was asfollows they wanted to create a
copycat product or series ofproducts this was in the fashion
space but not make it enoughpunitive as to really be
(10:25):
injuncted from court.
So they wanted to be cheekyenough to throw some bait and
since they were one of India'slargest multinational
corporations, they had a vastretail presence across several
Indian cities and online.
They wanted to leverage that toshow to the global brand that
(10:48):
they could do so much more ifthey were collaborating.
So, with discussion, jointly,we ended up.
Well, the company guys, thebrand guys, the business folks
ended up creating a productbased on our advice and launched
it.
We had a mega launch.
Naturally, it was meant to be abait throwing exercise and
(11:14):
therefore the bait was caughtand the large global brand sued
us.
And then we were successful indefending an infringement action
and didn't let them succeed ingetting an injunction against us
.
They went to the appellatebench and we succeeded again and
the matter got compromised andthe terms of compromise included
(11:39):
them handing over the globalexclusive distribution rights to
my client.
So we've had all sorts ofinteresting mandates over the
global exclusive distributionrights to my client.
So we've had all sorts ofinteresting mandates, but this
is one that just came to my mindbecause it's recent and it's
one of the more cheeky ones thatwe've got.
Speaker 2 (11:59):
Yes, congratulations.
Thank you, yeah well, I wishyou continued success with that.
Congratulations.
Thank you, yeah Well, I wishyou continued success with that.
Since you said, it is stillperhaps pending on appeal, is it
?
Speaker 1 (12:13):
Yes, yes, the trial is pending and it's probably now
, unfortunately, because inIndia matters take time to
resolve finally.
So it'll probably take a fewyears to finally get disposed of
, but we'll see how that goes.
Speaker 2 (12:30):
Yes, well, you've mentioned the Raghav project, so
perhaps you can give somebackground on that now.
Speaker 1 (12:39):
Before that, I just want to clarify the action that
they had brought for thatparticular product.
That one got compromised and,as terms of the compromise they
had brought for that particularproduct, that one got
compromised and, as terms of thecompromise, they withdrew the
suit and we we settled in inthose terms.
But there's a related actionfor another product which is
still pending, which we'recontesting, but on the raghav
(12:59):
aspect, um, yes, so thebackground to that is that
Kritika, my wife and I we werethese were early days of the
COVID, of the pandemic, march orApril, as far as my memory goes
2020.
And we were just, obviously,like the rest of the world,
stuck at home at home.
(13:29):
And well, almost since 2018,2019, we had been getting a lot
of queries from some of Europeanand US clients on various
aspects of automation,artificial intelligence, and if
such an element were to beintroduced in business, how
would it be treated under lawand in our context?
Obviously, the question wasdirected to us to give an answer
(13:51):
in terms of Indian law.
So you know, that gave us anidea as to the fact that this is
going to be, in the years tocome, one of the biggest
problems that the society willface one of the biggest
challenges because, as is alwaysthe case, technology evolves at
(14:14):
a much faster pace and law hasto lay catch up.
(14:41):
March we were walking and wethought of an idea that let's
just get a tool developed, we'reable to use the output as test
cases and can file someapplications in selected
jurisdictions that allow forcopyright to be applied for and
examined and registered.
So we shortlisted threecountries in the beginning.
(15:05):
We did India, we did US andCanada.
We also wanted to focus on UK,but unfortunately UK doesn't
have a system where you canapply for copyright protection
and be granted copyrightprotection in that sense.
So we decided to proceed withthese three jurisdictions.
(15:29):
We met a developer from Indianorigin who was doing his
master's.
He was an engineer from India.
He was doing his master's inthe Miller Institute, which is a
part of the University ofMontreal, specializing in
artificial intelligence.
His name was Raghav, and that'sthe story behind how the name
(15:52):
of the solution came to be Muchlater, of course.
So we approached them and saidlook, we have a small budget
which we're bootstrapping,kritika and myself, and we'd
really like you to develop thesolution.
So initially, him and a coupleof his colleagues who were
assisting him at the back.
They were of the view that theycould create something
(16:13):
pertaining to music, but then,eventually, we saw another one
of their solutions on a publicGitHub library which interested
us, and then we said look, guys,we're really impressed by your
neural style transfer solution.
(16:35):
Can you build something likethat for us?
And that's where we agreed.
And they built this neuralstyle transfer solution, and how
it works is very simple.
It takes two kinds of inputs ittakes a style input and it
takes a content input, and afterthat you have the option of
(16:59):
customizing what amount of stylegets transferred from the style
image to the content image anda couple of other variables, and
then you hit a button calledgenerate and then it produces an
output which is the tools,re-imagination of the style, the
(17:21):
content image in the styleimages, style, basically.
So how we set to go about it iswe wanted to minimize any
complications or any peripheralreasons for which IP offices
might just avoid answering themain question and raise an
objection to, for instance, whoowns the style image, who owns
(17:44):
the content image, and so on.
So, for simplicity sake, wedecided to keep for the content
image a picture that I clickedusing my phone like my regular
iPhone, and this was a time thatwe were walking on a terrace,
kritika and I, and it wasevening, it was sunset time, so
(18:05):
I just took my phone out andjust clicked a picture of the
sunset from my terrace, and thatis what we used as the base
image and the content image andfor style image.
To avoid any complication, wewanted to use a style that the
(18:26):
tool can be trained on.
So we chose Van Gogh, becauseall of Van Gogh's works are in
the public domain, naturally,and therefore it wouldn't have
led to any further complicationas regards the IP office in some
countries saying oh well, youguys don't own the rights to the
(18:50):
style image and therefore, onthat basis itself, we are going
to refuse registration.
So we chose an image and anartist which was popular.
They're painting an art styledistinctive, something that a
machine could detect, learn andmimic easily, and also in the
(19:13):
public domain.
So we narrowed down on one ofmy favorites, van Gogh, and we
chose the Starry Night because,simply because it worked for the
content image that we had for asunset scene, and then out came
the output.
I obviously tinkered aroundwith the variables, which was
(19:35):
how much style was to betransferred and so on, and then
the output that got generated iswhat was what we ended up
naming Suryasth, which is theHindi word for sunset, simply.
And then what we wanted to dowas file this as an image in
(19:58):
artistic work which was createdjointly with the effort of an AI
tool and a human being, andtherefore we decided to go the
co-author way in place of soleauthorship or any other
combination.
And the story behind naming itDrago is simple.
(20:19):
We said we don't know wherethis journey will lead to, but
at some stage, many years after,somebody or some government is
definitely going to ask thequestion what about the rights
of the developer who developedthe tool and trained the tool on
a set of images, and so on?
(20:40):
So he said look, you don'tfeature anywhere, so let's just
name the tool after you and callit raghav.
And then we created we kind ofdid some reverse engineering and
we created a name backwardsfrom the alphabets r-a-g-h-a-v,
so basically stands for robust,artificially intelligent
(21:03):
graphics and art visualizer.
So that's what it was.
Speaker 2 (21:11):
Very nice.
So this is circa 2020 that thisprocess starts and then, by
2021, you start sending in yourapplications to the three
jurisdictions that you've chosen, and so I guess I'm curious.
(21:32):
There are so many points I wantto follow up on with what you
just said, but at the same time,I'm curious about the very
interesting responses that youreceived from these
jurisdictions.
So would you like to give anoverview of that?
Speaker 1 (21:48):
Sure.
So you know, in fact, theprocess started 2020 itself.
We had, I think, nearly filedall of the applications by end
of 2020.
The first one to get registeredafter undergoing examination
was the Indian application, andwhat coincided with that was
that the Indian ParliamentaryStanding Committee on Commerce
(22:11):
was examining amendments orpotential amendments to the
existing intellectual propertylaw regime in view of emerging
technologies such as artificialintelligence, and I got invited
to depose as an expert witnessbefore the standing committee in
(22:33):
the Parliament House, and theyspecifically asked about my
endeavors and my application andwanted to understand what my
view was, as well as otherpeople's views were, on the
aspect of how to treat anAI-generated work under
(22:55):
copyright law or an AI-generatedinvention under patent law.
We already had the Davos matter.
It had taken place.
The patent one it definitelyhit the news by then in a big
way and there were questionsaround that as well.
So, to come back to each of thethree jurisdictions that we
(23:17):
filed in, so India was the firstto grant after considering our
submissions, and what we triedto argue to give you an outline
was as follows Basically, therewere a number of points on which
they objected to registering it, and one of them was the fact
(23:40):
that if the AI is to be made andconsidered as an author, it
will go against or contradictseveral parts of the act which
have been created with thepresumption that it has to be a
human being which will be anauthor.
(24:03):
So they drew our attention toseveral provisions.
Some of them included the factthat, for instance, the term of
copyright in case of jointauthorship gets calculated in
India from the death of thesubsequent, the second author,
and then the reason saying well,the AI is technically never
(24:23):
going to die, so how will theperiod be calculated?
To drive home the point thatthe law only required an AI to a
human to be an author, couldonly permit it a human to be an
author and non-human could notbe an author, and so on.
(24:44):
So how we tried to argue, thewhole point was as follows.
Of course there was also aquestion on whether they were
required to, or the work wassubject matter protection at all
or not, considering there wasconsiderable amount of computer
contribution to the work.
(25:05):
So first of all we drew theirattention to the fact that in
1994, indian law underwent someamendments which were based on
the mid to late 80s amendmentsof the UK law which gave
recognition tocomputer-generated works.
(25:26):
And the 94 amendment, althoughminor, made a significant impact
and change because theobjective of bringing that
amendment, amongst others, wasto give recognition to computer
generated works.
So what it specificallyprovided for was a clause which
(25:48):
sits under Section 2 of theIndian Corporate Act.
It's sort of like a deemingprovision.
Sectiond defines the termauthor and what that clause says
is for computer generated worksand there shall be the person
who causes the work to begenerated.
(26:10):
So by that logic they basicallygave recognition to the fact
that if there is a computergenerated work, so long as
there's an there's, there's anactual human who has given
commands or worked on thecomputer and so on and with
their use of the computeroperated it at the relevant time
(26:31):
when the work, the subject work, was created.
The person who's done that isentitled to claim authorship
over that work.
So we gave that logic and wesaid look, we understand.
There's several parts of theact that conflict with the
understanding that an AI cannotbe or a non-human cannot be an
(26:54):
author.
But there are two things thatyou should consider the current
form and the rules where theform is prescribed.
The delegated legislation, therules which the government's
created doesn't contain anyfield or column where I can
disclose these details, onadmission that it was in fact.
(27:15):
The subject work was in fact acreation of substantial input or
contribution by a computer.
It may have been a jointcreation, but it still had
substantial inputs from acomputer and therefore I had no
(27:35):
option but to, for attribution'ssake, name the application,
which is R Raghav, as aco-author, and what that meant
was I was ensuring that yourrecords, when you were to make
an entry in your records,correctly identify the fact that
(27:56):
this was created by me under2D6, with the assistance of
Raghav, which is a tool that Iown, and therefore I assist you,
or rather I fulfill myobligation of not concealing
material facts and you fulfillyour statutory obligation of
(28:18):
maintaining statutory recordswhich reflect the correct
factual position.
So that was argument number one.
The second argument was webasically fell back on a rather
technical principle ofinterpretation of statutes,
which was the Hayden's rule orthe mischief rule.
(28:38):
Was the Hayden's rule or themischief rule?
And how we use that to ouradvantage was by arguing that
look, if there are more than oneinterpretations possible, which
is happening in this case,because the law has been
designed, keeping intoconsideration that a non-human
cannot be an author.
But if there are more than oneinterpretations and there is a
(29:01):
conflict which is being created,it has to be resolved by taking
the interpretation whichfurthers the remedy and
suppresses the mischief.
In the context of the 1994amendment, we argued the
mischief that the amendmentsought to remedy or provide a
(29:22):
solution for was that there wasno express recognition under law
for computer-generated works.
So therefore, if the objectiveof that amendment was to give
recognition and thereby create apart to the protection of
computer-generated works, theinterpretation that enables me
to register this work,reflecting the correct factual
(29:45):
position, is actually theinterpretation which is
furthering the remedy that the1994 amendment sought to achieve
and suppressing the mischief,and therefore must be permitted.
So, on those arguments is howwe got our registration, and it
still sustains, funnily.
(30:05):
What is interesting is veryquickly to follow up on the
policy journey of that.
I obviously had a chance toexplain all of this to the
Parliamentary Standing Committeein 2020 December and then in
July 2021, they published theirrecommendation report, which I'm
(30:25):
happy to share.
I was very privileged to learnthat they did include several of
the recommendations that aremade, including the creation of
a separate category ofAI-generated works, possibly
with a lesser duration ofprotection or lesser scope of
(30:46):
protection, and the reasonbehind that was Stephanie
yourself.
More so because you're an expertin this field and deeply
associated with arts Myselfsimply as an admirer of arts and
someone having huge respect forartists and the talent that
they possess.
(31:06):
Kritika and I we've always beenof the view that, while it's
very important to embracetechnology, at the same time
rewarding human talent and inthis case rewarding human talent
and in this case, rewardinghuman ability to create art
always has to be kept at apedestal above something that
(31:29):
someone creates with theassistance of technology.
So taking the view thattechnology created works have to
be discarded altogether is alsonot correct in our humble view,
because that would lead topeople and the industry
discarding the use of emergingtechnology.
(31:50):
But, on the other hand,creating a fair balance would
entail that these rights arerecognized.
There's a formal procedure thatgovernments, that legislative
bodies, legislatures around theworld decide on in terms of how
someone who's created a workwith the assistance of an AI or,
(32:11):
in general, with the assistanceof a computer resource, has to
identify that on admission, andhow is that work to be treated
in terms of its duration ofprotection as well as its scope
of protection.
And what we suggested was wereminded them that, look, it's
not the first time somethinglike this would happen.
(32:32):
It happened when the camera wasinvented, it happened when
sound recordings incinematograph films were
introduced as works.
Each of these, in mostjurisdictions, enjoys much
shorter and a much more limitedscope of protection, simply
because it's the same logic.
These works are secondary worksor, in a certain sense,
(32:53):
derivative works, or assistedworks, or neighboring works
rising out of neighboring rights, and so on, because they rely
on substantial contribution,either from a device, such as in
the case of photography,obviously on cameras, which are
becoming smarter and smarterwith each day or they are
(33:14):
considered as having comprisedor based on several primary
categories of work, such as, forinstance, in the case of a song
.
It is made up of lyrical worksand musical works, and so on.
So we said look, you can createa sui generis system, you can
create a separate category ofrights.
(33:34):
And then there were a number ofquestions.
They asked us but how will itbe regulated?
How will it be adjudicated?
And then we put the questionback to them and we said how do
you think any artistic work, asimple artistic work, is
adjudicated today.
It's on faith and it's onself-declaration.
So a regulator, in any part ofthe country, any part of the
(33:56):
world, a corporate office, canonly investigate, uh to a
limited extent or examine to alimited extent.
Everything is based largely onself declarations.
So so long as the basicexamination is done, as per
procedure, and statements anddeclarations are found to be
accurate, registrations aregranted.
But everywhere registrationsare subject matter of challenge,
(34:20):
both in trial, in defense, tosay an infringement action, or
sometimes subject matter of acancellation action, as we're
facing in Canada.
So well, this was the Indianpart of it, which now brings me
to Canada.
So Canada, as far asregistration is concerned, was
very easy because they don'thave a formal examination and
(34:40):
they basically just go by whatyou submit and they kind of
create an entry in the recordsand they grant you a certificate
for it and then it's basicallya subject matter of either
proving or disproving that actin trial.
So what happened in Canada?
Is it simply proceeded to getregistered, although we were
(35:00):
rather surprised by the factthat we very clearly disclosed
again Raghav and the fact thatRaghav was a tool, a software,
an AI-based software veryclearly in the form itself
formal examination.
For some reason the Canadian IPoffice simply decided to
(35:21):
include it in the register ofcopyrights and grant us a
registration certificate.
So where that story is reachedis, as you may have read, we got
sued for cancellation ofregistration by CIPIC, which is
an IP policy clinic, and thatmatter is currently pending
(35:45):
before the Federal Court ofCanada and we're kind of in the
midst of proceedings.
From what I understand last,we're waiting for the written
submissions, written pleadings,of both sides to be completed
and then at some subsequentstage the matter will come up
for arguments before court.
So that's where the Canadianposition is, because SIPC,
obviously for a number ofreasons, believes that the
(36:07):
relevant entry for Sourios needsto be removed.
Alternatively, in the petitionwhich they filed, which they've
also made public on theirwebsite, as I've been told,
mentions an alternative plea oralternative approach to what
their prayer is.
Their ask is from court, whichis to remove Raghav as the
(36:30):
second author.
So they don't necessarilychallenge and even in their
submissions to the Canadiangovernment, in response to the
call for comments on thisspecific issue, the study that
the Canadian government hadcommissioned They've uploaded
their comments on their websiteand we are aware of the fact
that they've taken the positionthere that they're not
(36:51):
necessarily opposed tocomputer-generated or
AI-generated works not receivingprotection or receiving
protection, but they opposed tothe idea of any authorship or
similar rights flowing to anapplication.
So that's where the Canadianposition is.
The US position was verydifferent, in fact, and that's
(37:13):
where much of the actionhappened, because our US
application got examined andrefused sometime in 2021, after
which we sought the assistanceof a very renowned law firm,
particularly a very, verycompetent lawyer or team of
(37:35):
lawyers, which was led by a verycompetent lawyer called Alex
Garens at Dave Pitney, and sothe team at Dave Pitney, led by
Alex Garens, represented us inour first review and then in our
second review before the Boardof Review at the US Corporate
Office.
So the summary of what came outof the serious decision as is
(38:00):
published, also free foreveryone to view, on the US
corporate office website theyhave a specific page on AI and
it's one of the four decisionswhich they've published directly
there on that page.
What came out from thatdecision was following pointers
(38:20):
came out.
Number one they seem to besaying that well, we are not
going to give any copyrightprotection to anything which is
not human created.
Period Number two if there is awork which contains
contributions from an AI and ahuman being, so far as you can
(38:45):
discern what the human'scontribution was and what the
AI's contribution was, we willrecognize that, examine that and
then grant protection to theextent of the human being's
contribution and exclude fromthe purview protection the AI's
contribution totally.
And then, number three if theAI's contribution and human's
(39:09):
contribution is blended togetherin a manner that's impossible
to discern both elements, wewill not render any protection
to the work at all, whichhappened in our case and in the
Board of Review judgment.
They took an interesting turn.
They brought in the concept ofderivatives, which wasn't
(39:34):
previously either brought to ourattention or made a ground to
refuse protection.
So they said basically, as wesee it, the base image is what
is subject matter of copyright,because it's your photograph and
you can independently pursue acorporate protection for that.
(39:56):
And Sururya appears to be aderivative of that photograph
and, based on the submissionsand the facts that have been
brought to our attention, we areof the view that there has been
little or negligible humancontribution in making the
derivative work.
(40:17):
And they go to the extent ofsaying that the derivative work
is actually a product ofRaghav's contribution or
creative decisions and not MrSahni's contributions, because
they obviously apply the controltest and they say that
(40:37):
regardless of the fact Mr Sahniselected the base image, the
content image and the styleimage, and he also had an option
of deciding how much style totransfer.
The output really was veryremote from Mr Sani's control
and therefore it was impossiblefor the human author meaning
(41:01):
myself, meaning Ankit Saniactually being able to imagine
what the output would or haveany control over how the output
would look like.
And, amongst other arguments,the basic, most fundamental
argument that we made was lookwhat does the US Supreme Court
say?
Or what do you guys?
Do you guys protect images orphotographs that are generated
(41:28):
by these fancy DSLR cameraswhich, by the way, have far more
AI capabilities than Raghavdoes, capabilities than Raghav
does, and when you consider theeffort of clicking a shutter
button and selecting a monoexposure and zoom and so on, as
(41:49):
enough human creativity to beable to render full protection
to the photograph which iscreated digitally, absolutely
DSLR camera.
This goes beyond that, becauseI'm already working with one
picture which has been createdthe same way then I'm choosing,
(42:09):
making the creative decision ofchoosing Van Gogh's Starry Night
to be the relevant style forthis image, choosing how much
style to transfer to the contentimage, and then I'm generating
this image, choosing how muchstyle to transfer to the content
image, and then I'm generatingthis image.
But they somehow didn't agree onthat, which is possibly the
most bizarre thing to my mind,even today, and it seemed almost
(42:34):
as if the corporate office, orthe, should I say the
administration, the Democratadministration in place, it's
sort of philosophically orjurisprudentially arrived at the
conclusion that they're not byany means going to give
protection to machine generatedor AI generated works.
(42:55):
So they were trying to, at eachstep, somehow justify that
position by giving new reasons.
So this is an outline of whathas happened so far in each of
the three jurisdictions.
Speaker 2 (43:13):
Thank you for that.
You covered almost all thedetails that I was hoping that
you would, and so just a coupleof follow-ups For the current
definition in India of author.
So Section 2D6 has now beenclarified.
(43:35):
Has it changed per yourrecommendation?
Is that what you weresuggesting and has taken place?
Speaker 1 (43:43):
I just want to make sure I understood that correctly
created this sort of beamingfiction that in case of a
(44:04):
computer generated work, theperson who causes the work to be
generated can claim herself orhimself to be the author.
Now whether that will beconsidered as valid in case of
AI solutions.
And then there is a wholespectrum of variety of AI
solutions which differ in termsof how much input a human has to
(44:29):
give for it to generate anoutput.
Really, we have things likeMidjourney, popular tools like
Midjourney, etc.
And they obviously generate AItools and you basically, with a
few sentences and so on, youcould start having the tool
generate outputs and all of ushave done that many times at
(44:51):
some of the other points in ourlives.
But we have those kind of tools.
And then we have other kinds oftools in the market, as I
understand from my interactionwith clients, and tools in the
market, as I understand from myinteraction with clients and
others in the entertainmentmedia industries.
We've got tools like AutoTunethat work in a certain fashion
to correct a voice sample andput it in the right notes.
(45:15):
That is AI of a certain orderand a certain kind.
We've got special effects,specialized special effects
softwares that can populate agreen screen image of video with
placeholder content, whetherit's a waterfall or it's a scene
(45:35):
from mountains and so on, allabsolutely synthetically created
and relevant to what thedirector wants the scene to
really be.
So we have a wide variety of AItools and therefore I'm unsure
if and well, it doesn't matterwhat I think.
Actually, eventually it's amatter of how the regulator sees
(45:55):
it, the corporate office, orhow courts see it as to the
varying levels of human inputand effort that each of these
categories of tools require.
But as far as the section goes,it doesn't make any
qualifications.
It simply says anything that iscomputer generated, the person
that causes it to be generatedcan claim themselves to be
(46:17):
author, which leads to thepresumption that, therefore,
what the person caused to becreated or causes to be created
can be protected as copyright.
What has happened is thatdefinitions remain as it is.
We recommend it for a new kindof right to be amended into the
Corporate Act on the lines of,say, photograph or similar, into
(46:42):
the Corporate Act on the linesof, say, photograph or similar,
spelling out a new kind of act,maybe something like an
AI-assisted work category ofright and having a reduced
period of protection.
So the Standing Committee, theParliamentary Standing Committee
did recommend to the governmentthat they think that such a
(47:02):
right should be created.
This was 2021.
Interestingly, the last piece inthe puzzle so far is in
February of 2024, this year, theMinister of State the Union
Minister of State Comers inresponse to a question in the
Indian Parliament, made asubmission that a they don't
(47:28):
think there is any need tocreate a new category of right
and therefore the presentgovernment, the Indian
government, doesn't intend toamend the Act to create a new
category of right.
B they believe the way the actstands, the law stands today, it
is sufficiently well positionedto give protection to
(47:52):
AI-generated content and thegovernment intends to give
protection to AI-generatedcontent.
And that is where theybasically left it.
Of course, they did mention abit about, or implied a bit
about, text and data mining andthe fact that, in the absence of
(48:13):
any express exceptions, textand data mining efforts will be
seen against the test ofinfringement and the exceptions
which the statute provides.
And, should such text and datamining not be covered by any of
the exceptions, it may beconsidered as infringement if it
qualifies and fulfills allcriteria the definition of
(48:37):
infringement.
So that is the position inIndia as regards to this.
Speaker 2 (48:41):
Yeah, yeah, and my understanding for Canada is
there has been a recommendationreport that there should be
changes in that jurisdiction,but there hasn't been a timeline
put on it, and so there'sambiguity there and with your
case pending, who knows whichwill come out first clarity in
(49:03):
the law or your decision in yourcase?
Speaker 1 (49:07):
absolutely.
I mean, I'm really excited andthere's again a wonderful set of
attorneys who are representingus in canada, so I'm looking
forward to working with them andat some stage to being
physically present if I'mpermitted to in Canada before
the federal court, to be able toassist the court in making some
(49:29):
of these points and drivinghome some of our arguments.
Speaker 2 (49:35):
And then going then to the US.
I was curious your thoughtsabout some of the copyright
offices' request for commentsand did you actually submit any
formally to the copyright officewhen the window was open to do
it?
Speaker 1 (49:58):
I didn't even think till now to ask you that.
You know we took a consciouscall.
We did participate in somecountries' call for views.
Uk was one such country wherewe wanted to have our
submissions heard.
In the US I was invited to be aspeaker in the listening session
one of the listening sessions,one of the listening sessions
(50:19):
but Kritika and I we decided notto pursue or basically not to
submit our views because we feltmuch of what we needed to
submit had already been donethrough our submissions before
the corporate office and they'dalready made a decision on that,
and after the Board of Reviewdecision it would have been
improper to, in our view atleast, to really make any more
(50:41):
written submissions and so on.
So we basically decided to notparticipate.
But the fact that there is acall for views there is a rather
active subcommittee where I'maware, through certain people
who have either deposed beforethe subcommittee, the Senate
(51:03):
subcommittee, or have submittedtheir views in writing, the
Suryas case has been brought tothe attention of the Senate
subcommittee.
The fact that there is asubcommittee looking into it is
evidence by itself of the factthat the government, and
possibly the Senate, is ingeneral exploring whether the
(51:27):
law finally needs to be amendedin order to facilitate and make
room for artificial intelligenceand all of the other related
technologies that have takenover our lives.
Speaker 2 (51:38):
At this stage, yes, yes, and so for their specific
questions related to honing inon this conversation, the idea
of clarifying what theauthorship requirement is and
should be going forward and, asa policy matter, should that
change?
(51:58):
Does the Constitution allow forthat change?
All of those things, I think Ikind of already know your
thoughts, but if you'd want toput it into like a capsule, sure
?
Speaker 1 (52:10):
no, stephanie.
Thank you, and you know, in acapsule.
I do believe the authorshipunderstanding or the definition
should, by way of a legislativeamendment, be clarified or
broadened in scope to includeauthorship with the assistance
of technology for all categoriesof works, whether it's artistic
(52:34):
or cinematographic, or it'smusical, etc.
Or a cinematographic or it'smusical, et cetera.
Doing anything but that woulddiscourage the adoption of
technology by industries allacross the spectrum.
And if it doesn't discouragethe adoption of technology, it's
going to result in more andmore applicants suppressing
(52:54):
material facts, becauseotherwise, if you don't disclose
, there's no way that thecorporate office will actually
know if something was createdwith the assistance of
technology or not, and thereforethat that is what I feel is the
correct position.
I'm not any expert on usconstitutional law or us
(53:15):
corporate law, but from what Iunderstand, based on advice that
I've received from my attorneysin the past, I believe the
definition or the concept ofcopyright in the Constitution
and in the statute is reallyagnostic of whether the or
doesn't specifically clarifymention that an author has to be
(53:36):
a human.
Human authorship is a differentissue.
Debate for a different day.
Interestingly, we had ElonMusk's company announce Primus
or whatever Optimus, whateverthe name of the product is the
humanoid which you can buy andkeep at home for domestic
(53:58):
support and so on.
The fact is that with that,we'll have an AI which will be
walking, sooner than later, inour homes, in our offices and so
on.
That probably could be theright time to ask the question
as to whether a non-human can bean author or not, but so far as
the question the first questionis concerned, I think what is
(54:21):
important is for governments andcountries to take a view on
whether they are going toprotect works that are created
with the assistance of AI, andthat is the most important
question to answer at this stage.
Speaker 2 (54:34):
Yeah, and with all of the points that you've raised
today, do you have an opinionabout certain injustices that
might be percolating in ourcurrent global legal landscape,
and is there a need for a globalapproach to address those
(54:56):
injustices?
Speaker 1 (54:58):
to address those injustices.
You know, to me, when I seejustice again, you know I'm
probably not don't have enoughexperience to be able to comment
on something as broad andsomething as high level as this,
but you know, when I look at itI understand it simply, at the
very least, to mean that a partywho wishes to be heard is heard
(55:22):
in a fair and unbiased manner,and that, to me, is what justice
is.
Justice may not necessarilyresult in relief for the party
that brings the action, but sofar as the party gets an
unbiased consideration and ahearing of what their rights are
(55:45):
and what grievances they have,I think that fulfills the
definition of justice.
Of course, politically justicehas become rather complex and
unfortunately, in many parts ofthe world, including in India,
the judiciary has got mixed upwith politics and that doesn't
necessarily result in a goodcombination or a good cocktail,
(56:11):
if I could put it that way.
It's a dangerous cocktail toconsume and to have in one's
presence.
But in terms of injustices, inthe context of today's
conversation, I do feel thatcountries and governments should
work together, possibly at WIPOlevel or a similar forum level,
(56:34):
to create a treaty, a globaltreaty, to recognize and
therefore harmonize certainbasic, essential elements of how
AI and other emergingtechnologies and intellectual
property rights interact theinterface, and, once they agree
on a harmonized set ofprinciples, as they have been
(56:55):
doing since time immemorial forso many different kinds of
concepts and rights, whetherit's maritime law, it's
arbitration, things which I feelare more complex than the
interface of AI and IP Once theyhave that, then they could get
to legislating in theirrespective spaces their own
legislations.
But I feel governments shoulddo this sooner than later,
(57:19):
because what is happening isagain I refer to a conversation
with a colleague I respect a lot, who is based in London and is
somebody very senior in theentertainment media space, very
influential.
He told me that well, ankit, youknow how long do you think
we've been using AI tools inthis industry?
Since many years, possibly overa decade.
(57:41):
But in any of our copyrightpursuits, including
registrations of works, musicalworks or sound recordings we've
always, always suppressed thefact that it was created, at
times, with substantial machineassistance and inputs, with
substantial machine assistanceand inputs.
(58:02):
So that is what I feel is thefact which brings injustice the
most to people who are naturallytalented, because, at the end
of the day, the objective ofcorporate law or law in general
is to incentivize humans, is toprotect the rights of humans,
and to be putting somebody who'snaturally talented at the same
pedestal as someone who'screated something with the
(58:24):
assistance of a machine would beunfair, at least to my mind.
And that's happening today.
That kind of injustice ishappening because most places
people will suppress, largecompanies will suppress the fact
that they've used AI to createeither an invention or an
artistic work or so on, and thatneeds to be addressed and will
only be addressed if countriescome together to agree on a set
(58:49):
of harmonized principles andthen legislate and create a new
category and also createdelegated legislation.
The whole thing flows fromthere Examination guidelines and
rules and manner of disclosureand so on and so forth.
Speaker 2 (59:04):
Yeah, and the work you're doing hits the heart of
that, and so I wonder what isthe mark you hope to be creating
with this work that you'redoing?
Speaker 1 (59:18):
The same.
Actually, it might sound veryambitious, but when we sought to
do something like this in 2020,when we ventured out, and even
to date, the biggest motivationis to nudge IP offices, to nudge
governments and otherstakeholders into realizing that
(59:39):
the law needs to change.
And if our matter, if our caseor test cases, can make even a
small contribution to thechanges that are coming or will
come in the times that we'llwitness, I think I will consider
it as a rather satisfactoryoutput of all of the efforts
(01:00:02):
that we've made.
Of course, much of the creditgoes to all of the wonderful
people that inspired us alongthe way.
I have been immensely inspiredby Ryan Abbott and have often
fallen back on him for guidanceagain and again, and he has very
patiently not just guided mebut also helped me formulate
many legal arguments andpropositions.
(01:00:25):
And the same credit goes to thelaw firms that represented us.
A lot of the ideas came fromthem and then, of course, my
wife.
I mean she and I have beenideating this together.
So this is where we've reachedand that's what we hope we are
able to do, even if it'simpacting a very small part of
(01:00:45):
it.
Speaker 2 (01:00:48):
Yeah, and going forward.
Are there other projects likethis or any projects you want to
share that you're working on orplan to?
Speaker 1 (01:00:58):
So what we thought, and sometimes we discuss about
it, because much of what we'vedone was bootstrapped from
personal resources.
There's only a limit to howmuch you can really stretch a
legal experiment to.
But, on a light note, you know,whenever Pritika and I are
sitting over a breakfast tableon a Sunday and we sometimes
(01:01:19):
talk about where this debate hasreached, we always discuss and
agree that, look, if in the nextyear or couple of years we
don't see any results, it'llprobably be time for us to think
of something else, possiblycreating something with music
(01:01:39):
and AI, or creating somethingwith videography and then using
that to file test cases or drivehome at a different point, so
on.
Who knows?
But let's see where this effortgoes and then, if there are no
reasonable results, we mightjust end up taking the leap and
(01:02:01):
investing in or thinking ofsomething else as a solution.
Speaker 2 (01:02:06):
Yeah, well, I look forward to watching whatever it
is that you do, and thank you somuch for being here and sharing
your thoughts and the timelineof what you've been working on,
and is there anything else thatyou would wanna share or express
that we haven't touched on yet?
Speaker 1 (01:02:28):
Nothing specific, actually.
I mean, I'm really very happyto have conversed with you and I
would like to congratulate youfor taking forward many, many
important conversations aroundart and law, which will
hopefully not just shape or arealready shaping, as I understand
(01:02:50):
various issues and variousfacets of where art and law
intersect, but also lay thefoundation for more people to
brainstorm.
Whoever's listening to yourpodcast might just get inspired,
the same way we got inspired bysomeone else, something else,
and might take a step which willtake this debate and this
(01:03:12):
discussion to a whole differentlevel, and that is basically
what we hope for in the end.
Speaker 2 (01:03:17):
Thank you, and I certainly know that I and others
are inspired by what you'redoing, so thank you for that.
You know it occurs to mequickly.
I guess I am curious.
One final thought is how yousee this entire conversation and
the conflict that is kind ofemerging over these issues about
human authorship or how wedefine authorship.
(01:03:41):
How do you see that asimpacting our culture?
Speaker 1 (01:03:46):
You know, if you examine, in the earliest days,
when photography came intoexistence, a lot of naysayers I
like to refer to them asnaysayers because you've got
plenty of them in today's eraand age as well A lot of the
naysayers would.
And back then, as I readthrough several essays and works
(01:04:09):
on what happened at that pointin time, several naysayers would
say oh, you know, the artistsare going to have a run for
their money.
We would not have fine arts in20 years from now, because
photography is just so much morecapable of, for instance,
capturing an image.
And one of the essays that Iread, I imagine I remember
(01:04:31):
reading about the example of howthere are a number of artists
that line up and they've beenlining up for probably decades,
if not more hundreds of years,maybe in Paris, along River
Seine, and what they do is theymake customized in the moment,
(01:04:51):
literally they just createsketches of that scene, and
sometimes you'd have, dependingon what part of the river bank
they're on, what side they're on, sometimes you'd have the
Eiffel Tower in in front of themacross the river, sometimes you
would have the Louvre or otheryou know monuments, etc.
(01:05:12):
But so people back then wouldsay that look, all of these
folks that you see standingthere are not going to exist 20
years from now, because camerashave come and you will see a
bunch of camera photographersstanding here, would click your
pictures and give you the sameview, only with much more
clarity and precision that youcan carry back and in a fraction
(01:05:34):
of a time that it takes forsomeone to paint that photograph
.
But here we are in 2024,probably a couple of hundred
years, if not more, after theadvent of mainstream sort of
cameras in our lives, and Ithink the respect for art has
only grown and art has onlybecome more accessible to the
(01:05:57):
world than it was before.
So I feel, culturally, ai willhave much the same impact.
Where there's a use case forhumans to go up the value chain,
they will embrace AI as a toolto make their work better, their
outputs better.
(01:06:18):
But where there's a demand forand respect for natural talent,
I don't think, because if wewere to see history as evidence
of that, there should have beena decline in artistic talent in
general after the advent ofcameras, but that didn't happen.
So I have no reason topersonally believe that there
(01:06:39):
would be any decline in humantalent as far as music or art or
anything else is concerned.
On the contrary, there are adozen or more instances that one
can think of the top of thehead that AI would be useful in
in day-to-day work and in majorindustries, and so on.
Speaker 2 (01:06:57):
There will be links in the show notes to learn more.
If you were intrigued by thispodcast, it would be much
appreciated if you could leave arating or review and tag
Warfare of Art and Law podcast.
Until next time, this isStephanie Draughty bringing you
Warfare of Art and Law.
Thank you so much for listeningand remember injustice anywhere
(01:07:19):
is a threat to justiceeverywhere.
So I think we should embrace technology, as humans
have always had, withoutcarrying any apprehensions
regarding any negative thingsthat might impact culture or
society as a whole.
Speaker 2 (00:18):
Welcome to Warfare of Art and Law, the podcast that
focuses on how justice does ordoesn't play out when art and
law overlap.
Hi everyone, it's Stephanie,and that was intellectual
property and technology lawsattorney Ankit Sanhi.
In the following conversation,we discuss what drew him to IP
(00:40):
and tech law and the motivationbehind the Raghav Project and
tech law and the motivationbehind the Raghav project.
We also discuss the range ofresponses he's received from his
copyright applications to India, canada and the US for
registration of the Raghavoutput entitled Suriast, and we
(01:00):
close with considerations aboutthe definition of justice and
the cultural impact of emergingtech overlapping with art.
Ankit Sanhi, welcome to Warfareof Art and Law.
Thank you so much for beinghere.
Speaker 1 (01:19):
It's a pleasure, stephanie.
Thank you so much for giving methe opportunity to be on your
podcast, and I'm really lookingforward to our conversation
today.
Speaker 2 (01:28):
As am I, and perhaps we start with you giving a bit
about your background and yourwork.
Speaker 1 (01:36):
Sure.
So you know I've had a rathercliched background and career so
far as IP and technology lawsis concerned.
I'm a second generationtechnology and IP lawyer, so it
started very early, and almostin middle school is when I
(02:00):
nearly made up my mind that I'mgoing to go in the direction of
pursuing a law degree withspecialization in intellectual
property, innovation andtechnology.
I was always deeply passionateabout technology.
I was also my school's computersociety's president.
(02:23):
Back then this was theterminology that we would use
for anything which wastech-related.
So it was our school's computersociety, and I would basically
be very much engaged in allkinds of inter-school
competitions pertaining toprogramming and quizzing and so
(02:47):
on on various aspects oftechnology.
I also, incidentally, was oneof India's youngest to qualify
Microsoft certifiedprofessionals at the age of 14.
And so it kind of almost wasnatural for me to get an
(03:07):
opportunity to be where I am, asI'd always thought of being in
this kind of role well, that issuch a rich background.
Speaker 2 (03:20):
Like you are steeped in this work, and so I do wonder
, like you had opinions early onabout IP protection, can you
kind of walk us through that andhow they've evolved?
Speaker 1 (03:37):
Sure.
So by this do you mean how theconcept of IP protection may
have evolved?
Do you mean how?
Speaker 2 (03:44):
the concept of IP protection may have evolved.
Yeah well, just like youropinion.
Like some people are opensource of the opinion that open
source is the way to go, andothers are very steeped in
protecting and keeping it safefrom use.
So I wonder, like, has therebeen an arc in how you've seen
(04:07):
that or how is that for you?
Speaker 1 (04:11):
Yes, in fact, I'm so glad you brought up the question
.
You know, there's one thingthat I noticed in this journey,
and while we were growing up andyou know I went to law school,
and while we were growing up andI went to law school when we
start off as students, we'renaturally given various
perspectives, and the credit tothat goes to our teachers.
(04:35):
And what I've learned is, whenyou start out, the purpose of
really pursuing a law degree asyou may have felt yourself the
same way at some stage, Ibelieve is really to do justice
and your perspective's a bitmore leaning on the liberal side
(04:57):
than on the capitalist side,because that is the kind of
exposure, the academic exposure,that you get.
So, you know, one of the earlyworks that inspired me was Eric
Raymond's the Cathedral and theBazaar, and I was deeply
fascinated by the world of opensource.
(05:20):
Wikipedia had just become athing, wikipedia had just become
(05:49):
a thing, and for a model likethat to exist on global
contribution amongst a set ofpeople, individual parts that
they contributed, was thebiggest fascination.
But then, you know, the devilgets the best of us, as I like
to say.
And when you join theprofession, when you get the
opportunity to representcompanies, large corporations,
(06:09):
and defend their rights in courtor enforce their rights in
court of the spectrum all theway to pro-protection and
pro-right holders, becauseyou're really fighting for your
clients and representing themeach day in court or before the
(06:31):
IP office and so on.
Speaker 2 (06:38):
Yeah, absolutely, and I also.
Just as an aside, did you alsohave a background in the arts,
or like, what is your exposureand involvement with the arts?
Speaker 1 (06:54):
So I've always been passionate about art.
In general, I've had a thingfor theater and drama.
I used to be part of a theatergroup when I was in school, so
I've always had a lot ofadmiration for artists of our
era, artists of the times, theeras that have gone by, and in
(07:17):
that sense I remain associatedwith the arts.
But besides that, I haven'treally had any formal education
or training in the arts field assuch.
Speaker 2 (07:31):
Well, so then, I guess I would be curious to know
, like this, this law and techwork that you've been doing like
.
Are there certain cases thatexcuse me, certain cases that
stand out to you as beingemblematic of the work you do,
or certain ones you might wantto highlight?
Speaker 1 (07:57):
I've done more of cases on the trademark law side.
I'm primarily a trademarkenforcement person, if I could
put it that way.
Of course, we all have had ourfair share of contesting for
clients in design and copyrightcases as well.
I am fortunate to represent afew leading music artists and
(08:23):
some leading painters from India, and we represent their
interests both commercially aswell as in terms of protecting
their personality rights andother aspects.
But largely as a trademarklawyer, we've spent more time
defending large brands frombeing counterfeited or from bad
(08:49):
faith registrations or attemptsat passing off.
So I wouldn't say that myprofessional journey has really
had much association with theRaghav project or with Suryast
as such.
But you know there have alwaysbeen interesting matters.
I can probably spend a minuteor so talking about one of the
(09:15):
interesting mandates that we got, because it involves certain
aspects which are stillconfidential, to the extent that
that particular matter is stilllive.
I will refrain from using names, but I will try to fill in on
the interesting tidbits.
So we basically got a mandatefrom a large Indian
(09:38):
multinational corporation thatwas interested in, that was
interested in acquiring masterrights, licensee rights to a
large global brand, a retailbrand, and they came to us, the
business folks, saying that theytried their best to impress the
(10:00):
guys back in the United Statesfor giving them the exclusivity
for brand rights in India andthe subcontinent, but couldn't
succeed.
So their strategy was asfollows they wanted to create a
copycat product or series ofproducts this was in the fashion
space but not make it enoughpunitive as to really be
(10:25):
injuncted from court.
So they wanted to be cheekyenough to throw some bait and
since they were one of India'slargest multinational
corporations, they had a vastretail presence across several
Indian cities and online.
They wanted to leverage that toshow to the global brand that
(10:48):
they could do so much more ifthey were collaborating.
So, with discussion, jointly,we ended up.
Well, the company guys, thebrand guys, the business folks
ended up creating a productbased on our advice and launched
it.
We had a mega launch.
Naturally, it was meant to be abait throwing exercise and
(11:14):
therefore the bait was caughtand the large global brand sued
us.
And then we were successful indefending an infringement action
and didn't let them succeed ingetting an injunction against us
.
They went to the appellatebench and we succeeded again and
the matter got compromised andthe terms of compromise included
(11:39):
them handing over the globalexclusive distribution rights to
my client.
So we've had all sorts ofinteresting mandates over the
global exclusive distributionrights to my client.
So we've had all sorts ofinteresting mandates, but this
is one that just came to my mindbecause it's recent and it's
one of the more cheeky ones thatwe've got.
Speaker 2 (11:59):
Yes, congratulations.
Thank you, yeah well, I wishyou continued success with that.
Congratulations.
Thank you, yeah Well, I wishyou continued success with that.
Since you said, it is stillperhaps pending on appeal, is it
?
Speaker 1 (12:13):
Yes, yes, the trial is pending and it's probably now
, unfortunately, because inIndia matters take time to
resolve finally.
So it'll probably take a fewyears to finally get disposed of
, but we'll see how that goes.
Speaker 2 (12:30):
Yes, well, you've mentioned the Raghav project, so
perhaps you can give somebackground on that now.
Speaker 1 (12:39):
Before that, I just want to clarify the action that
they had brought for thatparticular product.
That one got compromised and,as terms of the compromise they
had brought for that particularproduct, that one got
compromised and, as terms of thecompromise, they withdrew the
suit and we we settled in inthose terms.
But there's a related actionfor another product which is
still pending, which we'recontesting, but on the raghav
(12:59):
aspect, um, yes, so thebackground to that is that
Kritika, my wife and I we werethese were early days of the
COVID, of the pandemic, march orApril, as far as my memory goes
2020.
And we were just, obviously,like the rest of the world,
stuck at home at home.
(13:29):
And well, almost since 2018,2019, we had been getting a lot
of queries from some of Europeanand US clients on various
aspects of automation,artificial intelligence, and if
such an element were to beintroduced in business, how
would it be treated under lawand in our context?
Obviously, the question wasdirected to us to give an answer
(13:51):
in terms of Indian law.
So you know, that gave us anidea as to the fact that this is
going to be, in the years tocome, one of the biggest
problems that the society willface one of the biggest
challenges because, as is alwaysthe case, technology evolves at
(14:14):
a much faster pace and law hasto lay catch up.
(14:41):
March we were walking and wethought of an idea that let's
just get a tool developed, we'reable to use the output as test
cases and can file someapplications in selected
jurisdictions that allow forcopyright to be applied for and
examined and registered.
So we shortlisted threecountries in the beginning.
(15:05):
We did India, we did US andCanada.
We also wanted to focus on UK,but unfortunately UK doesn't
have a system where you canapply for copyright protection
and be granted copyrightprotection in that sense.
So we decided to proceed withthese three jurisdictions.
(15:29):
We met a developer from Indianorigin who was doing his
master's.
He was an engineer from India.
He was doing his master's inthe Miller Institute, which is a
part of the University ofMontreal, specializing in
artificial intelligence.
His name was Raghav, and that'sthe story behind how the name
(15:52):
of the solution came to be Muchlater, of course.
So we approached them and saidlook, we have a small budget
which we're bootstrapping,kritika and myself, and we'd
really like you to develop thesolution.
So initially, him and a coupleof his colleagues who were
assisting him at the back.
They were of the view that theycould create something
(16:13):
pertaining to music, but then,eventually, we saw another one
of their solutions on a publicGitHub library which interested
us, and then we said look, guys,we're really impressed by your
neural style transfer solution.
(16:35):
Can you build something likethat for us?
And that's where we agreed.
And they built this neuralstyle transfer solution, and how
it works is very simple.
It takes two kinds of inputs ittakes a style input and it
takes a content input, and afterthat you have the option of
(16:59):
customizing what amount of stylegets transferred from the style
image to the content image anda couple of other variables, and
then you hit a button calledgenerate and then it produces an
output which is the tools,re-imagination of the style, the
(17:21):
content image in the styleimages, style, basically.
So how we set to go about it iswe wanted to minimize any
complications or any peripheralreasons for which IP offices
might just avoid answering themain question and raise an
objection to, for instance, whoowns the style image, who owns
(17:44):
the content image, and so on.
So, for simplicity sake, wedecided to keep for the content
image a picture that I clickedusing my phone like my regular
iPhone, and this was a time thatwe were walking on a terrace,
kritika and I, and it wasevening, it was sunset time, so
(18:05):
I just took my phone out andjust clicked a picture of the
sunset from my terrace, and thatis what we used as the base
image and the content image andfor style image.
To avoid any complication, wewanted to use a style that the
(18:26):
tool can be trained on.
So we chose Van Gogh, becauseall of Van Gogh's works are in
the public domain, naturally,and therefore it wouldn't have
led to any further complicationas regards the IP office in some
countries saying oh well, youguys don't own the rights to the
(18:50):
style image and therefore, onthat basis itself, we are going
to refuse registration.
So we chose an image and anartist which was popular.
They're painting an art styledistinctive, something that a
machine could detect, learn andmimic easily, and also in the
(19:13):
public domain.
So we narrowed down on one ofmy favorites, van Gogh, and we
chose the Starry Night because,simply because it worked for the
content image that we had for asunset scene, and then out came
the output.
I obviously tinkered aroundwith the variables, which was
(19:35):
how much style was to betransferred and so on, and then
the output that got generated iswhat was what we ended up
naming Suryasth, which is theHindi word for sunset, simply.
And then what we wanted to dowas file this as an image in
(19:58):
artistic work which was createdjointly with the effort of an AI
tool and a human being, andtherefore we decided to go the
co-author way in place of soleauthorship or any other
combination.
And the story behind naming itDrago is simple.
(20:19):
We said we don't know wherethis journey will lead to, but
at some stage, many years after,somebody or some government is
definitely going to ask thequestion what about the rights
of the developer who developedthe tool and trained the tool on
a set of images, and so on?
(20:40):
So he said look, you don'tfeature anywhere, so let's just
name the tool after you and callit raghav.
And then we created we kind ofdid some reverse engineering and
we created a name backwardsfrom the alphabets r-a-g-h-a-v,
so basically stands for robust,artificially intelligent
(21:03):
graphics and art visualizer.
So that's what it was.
Speaker 2 (21:11):
Very nice.
So this is circa 2020 that thisprocess starts and then, by
2021, you start sending in yourapplications to the three
jurisdictions that you've chosen, and so I guess I'm curious.
(21:32):
There are so many points I wantto follow up on with what you
just said, but at the same time,I'm curious about the very
interesting responses that youreceived from these
jurisdictions.
So would you like to give anoverview of that?
Speaker 1 (21:48):
Sure.
So you know, in fact, theprocess started 2020 itself.
We had, I think, nearly filedall of the applications by end
of 2020.
The first one to get registeredafter undergoing examination
was the Indian application, andwhat coincided with that was
that the Indian ParliamentaryStanding Committee on Commerce
(22:11):
was examining amendments orpotential amendments to the
existing intellectual propertylaw regime in view of emerging
technologies such as artificialintelligence, and I got invited
to depose as an expert witnessbefore the standing committee in
(22:33):
the Parliament House, and theyspecifically asked about my
endeavors and my application andwanted to understand what my
view was, as well as otherpeople's views were, on the
aspect of how to treat anAI-generated work under
(22:55):
copyright law or an AI-generatedinvention under patent law.
We already had the Davos matter.
It had taken place.
The patent one it definitelyhit the news by then in a big
way and there were questionsaround that as well.
So, to come back to each of thethree jurisdictions that we
(23:17):
filed in, so India was the firstto grant after considering our
submissions, and what we triedto argue to give you an outline
was as follows Basically, therewere a number of points on which
they objected to registering it, and one of them was the fact
(23:40):
that if the AI is to be made andconsidered as an author, it
will go against or contradictseveral parts of the act which
have been created with thepresumption that it has to be a
human being which will be anauthor.
(24:03):
So they drew our attention toseveral provisions.
Some of them included the factthat, for instance, the term of
copyright in case of jointauthorship gets calculated in
India from the death of thesubsequent, the second author,
and then the reason saying well,the AI is technically never
(24:23):
going to die, so how will theperiod be calculated?
To drive home the point thatthe law only required an AI to a
human to be an author, couldonly permit it a human to be an
author and non-human could notbe an author, and so on.
(24:44):
So how we tried to argue, thewhole point was as follows.
Of course there was also aquestion on whether they were
required to, or the work wassubject matter protection at all
or not, considering there wasconsiderable amount of computer
contribution to the work.
(25:05):
So first of all we drew theirattention to the fact that in
1994, indian law underwent someamendments which were based on
the mid to late 80s amendmentsof the UK law which gave
recognition tocomputer-generated works.
(25:26):
And the 94 amendment, althoughminor, made a significant impact
and change because theobjective of bringing that
amendment, amongst others, wasto give recognition to computer
generated works.
So what it specificallyprovided for was a clause which
(25:48):
sits under Section 2 of theIndian Corporate Act.
It's sort of like a deemingprovision.
Sectiond defines the termauthor and what that clause says
is for computer generated worksand there shall be the person
who causes the work to begenerated.
(26:10):
So by that logic they basicallygave recognition to the fact
that if there is a computergenerated work, so long as
there's an there's, there's anactual human who has given
commands or worked on thecomputer and so on and with
their use of the computeroperated it at the relevant time
(26:31):
when the work, the subject work, was created.
The person who's done that isentitled to claim authorship
over that work.
So we gave that logic and wesaid look, we understand.
There's several parts of theact that conflict with the
understanding that an AI cannotbe or a non-human cannot be an
(26:54):
author.
But there are two things thatyou should consider the current
form and the rules where theform is prescribed.
The delegated legislation, therules which the government's
created doesn't contain anyfield or column where I can
disclose these details, onadmission that it was in fact.
(27:15):
The subject work was in fact acreation of substantial input or
contribution by a computer.
It may have been a jointcreation, but it still had
substantial inputs from acomputer and therefore I had no
(27:35):
option but to, for attribution'ssake, name the application,
which is R Raghav, as aco-author, and what that meant
was I was ensuring that yourrecords, when you were to make
an entry in your records,correctly identify the fact that
(27:56):
this was created by me under2D6, with the assistance of
Raghav, which is a tool that Iown, and therefore I assist you,
or rather I fulfill myobligation of not concealing
material facts and you fulfillyour statutory obligation of
(28:18):
maintaining statutory recordswhich reflect the correct
factual position.
So that was argument number one.
The second argument was webasically fell back on a rather
technical principle ofinterpretation of statutes,
which was the Hayden's rule orthe mischief rule.
(28:38):
Was the Hayden's rule or themischief rule?
And how we use that to ouradvantage was by arguing that
look, if there are more than oneinterpretations possible, which
is happening in this case,because the law has been
designed, keeping intoconsideration that a non-human
cannot be an author.
But if there are more than oneinterpretations and there is a
(29:01):
conflict which is being created,it has to be resolved by taking
the interpretation whichfurthers the remedy and
suppresses the mischief.
In the context of the 1994amendment, we argued the
mischief that the amendmentsought to remedy or provide a
(29:22):
solution for was that there wasno express recognition under law
for computer-generated works.
So therefore, if the objectiveof that amendment was to give
recognition and thereby create apart to the protection of
computer-generated works, theinterpretation that enables me
to register this work,reflecting the correct factual
(29:45):
position, is actually theinterpretation which is
furthering the remedy that the1994 amendment sought to achieve
and suppressing the mischief,and therefore must be permitted.
So, on those arguments is howwe got our registration, and it
still sustains, funnily.
(30:05):
What is interesting is veryquickly to follow up on the
policy journey of that.
I obviously had a chance toexplain all of this to the
Parliamentary Standing Committeein 2020 December and then in
July 2021, they published theirrecommendation report, which I'm
(30:25):
happy to share.
I was very privileged to learnthat they did include several of
the recommendations that aremade, including the creation of
a separate category ofAI-generated works, possibly
with a lesser duration ofprotection or lesser scope of
(30:46):
protection, and the reasonbehind that was Stephanie
yourself.
More so because you're an expertin this field and deeply
associated with arts Myselfsimply as an admirer of arts and
someone having huge respect forartists and the talent that
they possess.
(31:06):
Kritika and I we've always beenof the view that, while it's
very important to embracetechnology, at the same time
rewarding human talent and inthis case rewarding human talent
and in this case, rewardinghuman ability to create art
always has to be kept at apedestal above something that
(31:29):
someone creates with theassistance of technology.
So taking the view thattechnology created works have to
be discarded altogether is alsonot correct in our humble view,
because that would lead topeople and the industry
discarding the use of emergingtechnology.
(31:50):
But, on the other hand,creating a fair balance would
entail that these rights arerecognized.
There's a formal procedure thatgovernments, that legislative
bodies, legislatures around theworld decide on in terms of how
someone who's created a workwith the assistance of an AI or,
(32:11):
in general, with the assistanceof a computer resource, has to
identify that on admission, andhow is that work to be treated
in terms of its duration ofprotection as well as its scope
of protection.
And what we suggested was wereminded them that, look, it's
not the first time somethinglike this would happen.
(32:32):
It happened when the camera wasinvented, it happened when
sound recordings incinematograph films were
introduced as works.
Each of these, in mostjurisdictions, enjoys much
shorter and a much more limitedscope of protection, simply
because it's the same logic.
These works are secondary worksor, in a certain sense,
(32:53):
derivative works, or assistedworks, or neighboring works
rising out of neighboring rights, and so on, because they rely
on substantial contribution,either from a device, such as in
the case of photography,obviously on cameras, which are
becoming smarter and smarterwith each day or they are
(33:14):
considered as having comprisedor based on several primary
categories of work, such as, forinstance, in the case of a song
.
It is made up of lyrical worksand musical works, and so on.
So we said look, you can createa sui generis system, you can
create a separate category ofrights.
(33:34):
And then there were a number ofquestions.
They asked us but how will itbe regulated?
How will it be adjudicated?
And then we put the questionback to them and we said how do
you think any artistic work, asimple artistic work, is
adjudicated today.
It's on faith and it's onself-declaration.
So a regulator, in any part ofthe country, any part of the
(33:56):
world, a corporate office, canonly investigate, uh to a
limited extent or examine to alimited extent.
Everything is based largely onself declarations.
So so long as the basicexamination is done, as per
procedure, and statements anddeclarations are found to be
accurate, registrations aregranted.
But everywhere registrationsare subject matter of challenge,
(34:20):
both in trial, in defense, tosay an infringement action, or
sometimes subject matter of acancellation action, as we're
facing in Canada.
So well, this was the Indianpart of it, which now brings me
to Canada.
So Canada, as far asregistration is concerned, was
very easy because they don'thave a formal examination and
(34:40):
they basically just go by whatyou submit and they kind of
create an entry in the recordsand they grant you a certificate
for it and then it's basicallya subject matter of either
proving or disproving that actin trial.
So what happened in Canada?
Is it simply proceeded to getregistered, although we were
(35:00):
rather surprised by the factthat we very clearly disclosed
again Raghav and the fact thatRaghav was a tool, a software,
an AI-based software veryclearly in the form itself
formal examination.
For some reason the Canadian IPoffice simply decided to
(35:21):
include it in the register ofcopyrights and grant us a
registration certificate.
So where that story is reachedis, as you may have read, we got
sued for cancellation ofregistration by CIPIC, which is
an IP policy clinic, and thatmatter is currently pending
(35:45):
before the Federal Court ofCanada and we're kind of in the
midst of proceedings.
From what I understand last,we're waiting for the written
submissions, written pleadings,of both sides to be completed
and then at some subsequentstage the matter will come up
for arguments before court.
So that's where the Canadianposition is, because SIPC,
obviously for a number ofreasons, believes that the
(36:07):
relevant entry for Sourios needsto be removed.
Alternatively, in the petitionwhich they filed, which they've
also made public on theirwebsite, as I've been told,
mentions an alternative plea oralternative approach to what
their prayer is.
Their ask is from court, whichis to remove Raghav as the
(36:30):
second author.
So they don't necessarilychallenge and even in their
submissions to the Canadiangovernment, in response to the
call for comments on thisspecific issue, the study that
the Canadian government hadcommissioned They've uploaded
their comments on their websiteand we are aware of the fact
that they've taken the positionthere that they're not
(36:51):
necessarily opposed tocomputer-generated or
AI-generated works not receivingprotection or receiving
protection, but they opposed tothe idea of any authorship or
similar rights flowing to anapplication.
So that's where the Canadianposition is.
The US position was verydifferent, in fact, and that's
(37:13):
where much of the actionhappened, because our US
application got examined andrefused sometime in 2021, after
which we sought the assistanceof a very renowned law firm,
particularly a very, verycompetent lawyer or team of
(37:35):
lawyers, which was led by a verycompetent lawyer called Alex
Garens at Dave Pitney, and sothe team at Dave Pitney, led by
Alex Garens, represented us inour first review and then in our
second review before the Boardof Review at the US Corporate
Office.
So the summary of what came outof the serious decision as is
(38:00):
published, also free foreveryone to view, on the US
corporate office website theyhave a specific page on AI and
it's one of the four decisionswhich they've published directly
there on that page.
What came out from thatdecision was following pointers
(38:20):
came out.
Number one they seem to besaying that well, we are not
going to give any copyrightprotection to anything which is
not human created.
Period Number two if there is awork which contains
contributions from an AI and ahuman being, so far as you can
(38:45):
discern what the human'scontribution was and what the
AI's contribution was, we willrecognize that, examine that and
then grant protection to theextent of the human being's
contribution and exclude fromthe purview protection the AI's
contribution totally.
And then, number three if theAI's contribution and human's
(39:09):
contribution is blended togetherin a manner that's impossible
to discern both elements, wewill not render any protection
to the work at all, whichhappened in our case and in the
Board of Review judgment.
They took an interesting turn.
They brought in the concept ofderivatives, which wasn't
(39:34):
previously either brought to ourattention or made a ground to
refuse protection.
So they said basically, as wesee it, the base image is what
is subject matter of copyright,because it's your photograph and
you can independently pursue acorporate protection for that.
(39:56):
And Sururya appears to be aderivative of that photograph
and, based on the submissionsand the facts that have been
brought to our attention, we areof the view that there has been
little or negligible humancontribution in making the
derivative work.
(40:17):
And they go to the extent ofsaying that the derivative work
is actually a product ofRaghav's contribution or
creative decisions and not MrSahni's contributions, because
they obviously apply the controltest and they say that
(40:37):
regardless of the fact Mr Sahniselected the base image, the
content image and the styleimage, and he also had an option
of deciding how much style totransfer.
The output really was veryremote from Mr Sani's control
and therefore it was impossiblefor the human author meaning
(41:01):
myself, meaning Ankit Saniactually being able to imagine
what the output would or haveany control over how the output
would look like.
And, amongst other arguments,the basic, most fundamental
argument that we made was lookwhat does the US Supreme Court
say?
Or what do you guys?
Do you guys protect images orphotographs that are generated
(41:28):
by these fancy DSLR cameraswhich, by the way, have far more
AI capabilities than Raghavdoes, capabilities than Raghav
does, and when you consider theeffort of clicking a shutter
button and selecting a monoexposure and zoom and so on, as
(41:49):
enough human creativity to beable to render full protection
to the photograph which iscreated digitally, absolutely
DSLR camera.
This goes beyond that, becauseI'm already working with one
picture which has been createdthe same way then I'm choosing,
(42:09):
making the creative decision ofchoosing Van Gogh's Starry Night
to be the relevant style forthis image, choosing how much
style to transfer to the contentimage, and then I'm generating
this image, choosing how muchstyle to transfer to the content
image, and then I'm generatingthis image.
But they somehow didn't agree onthat, which is possibly the
most bizarre thing to my mind,even today, and it seemed almost
(42:34):
as if the corporate office, orthe, should I say the
administration, the Democratadministration in place, it's
sort of philosophically orjurisprudentially arrived at the
conclusion that they're not byany means going to give
protection to machine generatedor AI generated works.
(42:55):
So they were trying to, at eachstep, somehow justify that
position by giving new reasons.
So this is an outline of whathas happened so far in each of
the three jurisdictions.
Speaker 2 (43:13):
Thank you for that.
You covered almost all thedetails that I was hoping that
you would, and so just a coupleof follow-ups For the current
definition in India of author.
So Section 2D6 has now beenclarified.
(43:35):
Has it changed per yourrecommendation?
Is that what you weresuggesting and has taken place?
Speaker 1 (43:43):
I just want to make sure I understood that correctly
created this sort of beamingfiction that in case of a
(44:04):
computer generated work, theperson who causes the work to be
generated can claim herself orhimself to be the author.
Now whether that will beconsidered as valid in case of
AI solutions.
And then there is a wholespectrum of variety of AI
solutions which differ in termsof how much input a human has to
(44:29):
give for it to generate anoutput.
Really, we have things likeMidjourney, popular tools like
Midjourney, etc.
And they obviously generate AItools and you basically, with a
few sentences and so on, youcould start having the tool
generate outputs and all of ushave done that many times at
(44:51):
some of the other points in ourlives.
But we have those kind of tools.
And then we have other kinds oftools in the market, as I
understand from my interactionwith clients, and tools in the
market, as I understand from myinteraction with clients and
others in the entertainmentmedia industries.
We've got tools like AutoTunethat work in a certain fashion
to correct a voice sample andput it in the right notes.
(45:15):
That is AI of a certain orderand a certain kind.
We've got special effects,specialized special effects
softwares that can populate agreen screen image of video with
placeholder content, whetherit's a waterfall or it's a scene
(45:35):
from mountains and so on, allabsolutely synthetically created
and relevant to what thedirector wants the scene to
really be.
So we have a wide variety of AItools and therefore I'm unsure
if and well, it doesn't matterwhat I think.
Actually, eventually it's amatter of how the regulator sees
(45:55):
it, the corporate office, orhow courts see it as to the
varying levels of human inputand effort that each of these
categories of tools require.
But as far as the section goes,it doesn't make any
qualifications.
It simply says anything that iscomputer generated, the person
that causes it to be generatedcan claim themselves to be
(46:17):
author, which leads to thepresumption that, therefore,
what the person caused to becreated or causes to be created
can be protected as copyright.
What has happened is thatdefinitions remain as it is.
We recommend it for a new kindof right to be amended into the
Corporate Act on the lines of,say, photograph or similar, into
(46:42):
the Corporate Act on the linesof, say, photograph or similar,
spelling out a new kind of act,maybe something like an
AI-assisted work category ofright and having a reduced
period of protection.
So the Standing Committee, theParliamentary Standing Committee
did recommend to the governmentthat they think that such a
(47:02):
right should be created.
This was 2021.
Interestingly, the last piece inthe puzzle so far is in
February of 2024, this year, theMinister of State the Union
Minister of State Comers inresponse to a question in the
Indian Parliament, made asubmission that a they don't
(47:28):
think there is any need tocreate a new category of right
and therefore the presentgovernment, the Indian
government, doesn't intend toamend the Act to create a new
category of right.
B they believe the way the actstands, the law stands today, it
is sufficiently well positionedto give protection to
(47:52):
AI-generated content and thegovernment intends to give
protection to AI-generatedcontent.
And that is where theybasically left it.
Of course, they did mention abit about, or implied a bit
about, text and data mining andthe fact that, in the absence of
(48:13):
any express exceptions, textand data mining efforts will be
seen against the test ofinfringement and the exceptions
which the statute provides.
And, should such text and datamining not be covered by any of
the exceptions, it may beconsidered as infringement if it
qualifies and fulfills allcriteria the definition of
(48:37):
infringement.
So that is the position inIndia as regards to this.
Speaker 2 (48:41):
Yeah, yeah, and my understanding for Canada is
there has been a recommendationreport that there should be
changes in that jurisdiction,but there hasn't been a timeline
put on it, and so there'sambiguity there and with your
case pending, who knows whichwill come out first clarity in
(49:03):
the law or your decision in yourcase?
Speaker 1 (49:07):
absolutely.
I mean, I'm really excited andthere's again a wonderful set of
attorneys who are representingus in canada, so I'm looking
forward to working with them andat some stage to being
physically present if I'mpermitted to in Canada before
the federal court, to be able toassist the court in making some
(49:29):
of these points and drivinghome some of our arguments.
Speaker 2 (49:35):
And then going then to the US.
I was curious your thoughtsabout some of the copyright
offices' request for commentsand did you actually submit any
formally to the copyright officewhen the window was open to do
it?
Speaker 1 (49:58):
I didn't even think till now to ask you that.
You know we took a consciouscall.
We did participate in somecountries' call for views.
Uk was one such country wherewe wanted to have our
submissions heard.
In the US I was invited to be aspeaker in the listening session
one of the listening sessions,one of the listening sessions
(50:19):
but Kritika and I we decided notto pursue or basically not to
submit our views because we feltmuch of what we needed to
submit had already been donethrough our submissions before
the corporate office and they'dalready made a decision on that,
and after the Board of Reviewdecision it would have been
improper to, in our view atleast, to really make any more
(50:41):
written submissions and so on.
So we basically decided to notparticipate.
But the fact that there is acall for views there is a rather
active subcommittee where I'maware, through certain people
who have either deposed beforethe subcommittee, the Senate
(51:03):
subcommittee, or have submittedtheir views in writing, the
Suryas case has been brought tothe attention of the Senate
subcommittee.
The fact that there is asubcommittee looking into it is
evidence by itself of the factthat the government, and
possibly the Senate, is ingeneral exploring whether the
(51:27):
law finally needs to be amendedin order to facilitate and make
room for artificial intelligenceand all of the other related
technologies that have takenover our lives.
Speaker 2 (51:38):
At this stage, yes, yes, and so for their specific
questions related to honing inon this conversation, the idea
of clarifying what theauthorship requirement is and
should be going forward and, asa policy matter, should that
change?
(51:58):
Does the Constitution allow forthat change?
All of those things, I think Ikind of already know your
thoughts, but if you'd want toput it into like a capsule, sure
?
Speaker 1 (52:10):
no, stephanie.
Thank you, and you know, in acapsule.
I do believe the authorshipunderstanding or the definition
should, by way of a legislativeamendment, be clarified or
broadened in scope to includeauthorship with the assistance
of technology for all categoriesof works, whether it's artistic
(52:34):
or cinematographic, or it'smusical, etc.
Or a cinematographic or it'smusical, et cetera.
Doing anything but that woulddiscourage the adoption of
technology by industries allacross the spectrum.
And if it doesn't discouragethe adoption of technology, it's
going to result in more andmore applicants suppressing
(52:54):
material facts, becauseotherwise, if you don't disclose
, there's no way that thecorporate office will actually
know if something was createdwith the assistance of
technology or not, and thereforethat that is what I feel is the
correct position.
I'm not any expert on usconstitutional law or us
(53:15):
corporate law, but from what Iunderstand, based on advice that
I've received from my attorneysin the past, I believe the
definition or the concept ofcopyright in the Constitution
and in the statute is reallyagnostic of whether the or
doesn't specifically clarifymention that an author has to be
(53:36):
a human.
Human authorship is a differentissue.
Debate for a different day.
Interestingly, we had ElonMusk's company announce Primus
or whatever Optimus, whateverthe name of the product is the
humanoid which you can buy andkeep at home for domestic
(53:58):
support and so on.
The fact is that with that,we'll have an AI which will be
walking, sooner than later, inour homes, in our offices and so
on.
That probably could be theright time to ask the question
as to whether a non-human can bean author or not, but so far as
the question the first questionis concerned, I think what is
(54:21):
important is for governments andcountries to take a view on
whether they are going toprotect works that are created
with the assistance of AI, andthat is the most important
question to answer at this stage.
Speaker 2 (54:34):
Yeah, and with all of the points that you've raised
today, do you have an opinionabout certain injustices that
might be percolating in ourcurrent global legal landscape,
and is there a need for a globalapproach to address those
(54:56):
injustices?
Speaker 1 (54:58):
to address those injustices.
You know, to me, when I seejustice again, you know I'm
probably not don't have enoughexperience to be able to comment
on something as broad andsomething as high level as this,
but you know, when I look at itI understand it simply, at the
very least, to mean that a partywho wishes to be heard is heard
(55:22):
in a fair and unbiased manner,and that, to me, is what justice
is.
Justice may not necessarilyresult in relief for the party
that brings the action, but sofar as the party gets an
unbiased consideration and ahearing of what their rights are
(55:45):
and what grievances they have,I think that fulfills the
definition of justice.
Of course, politically justicehas become rather complex and
unfortunately, in many parts ofthe world, including in India,
the judiciary has got mixed upwith politics and that doesn't
necessarily result in a goodcombination or a good cocktail,
(56:11):
if I could put it that way.
It's a dangerous cocktail toconsume and to have in one's
presence.
But in terms of injustices, inthe context of today's
conversation, I do feel thatcountries and governments should
work together, possibly at WIPOlevel or a similar forum level,
(56:34):
to create a treaty, a globaltreaty, to recognize and
therefore harmonize certainbasic, essential elements of how
AI and other emergingtechnologies and intellectual
property rights interact theinterface, and, once they agree
on a harmonized set ofprinciples, as they have been
(56:55):
doing since time immemorial forso many different kinds of
concepts and rights, whetherit's maritime law, it's
arbitration, things which I feelare more complex than the
interface of AI and IP Once theyhave that, then they could get
to legislating in theirrespective spaces their own
legislations.
But I feel governments shoulddo this sooner than later,
(57:19):
because what is happening isagain I refer to a conversation
with a colleague I respect a lot, who is based in London and is
somebody very senior in theentertainment media space, very
influential.
He told me that well, ankit, youknow how long do you think
we've been using AI tools inthis industry?
Since many years, possibly overa decade.
(57:41):
But in any of our copyrightpursuits, including
registrations of works, musicalworks or sound recordings we've
always, always suppressed thefact that it was created, at
times, with substantial machineassistance and inputs, with
substantial machine assistanceand inputs.
(58:02):
So that is what I feel is thefact which brings injustice the
most to people who are naturallytalented, because, at the end
of the day, the objective ofcorporate law or law in general
is to incentivize humans, is toprotect the rights of humans,
and to be putting somebody who'snaturally talented at the same
pedestal as someone who'screated something with the
(58:24):
assistance of a machine would beunfair, at least to my mind.
And that's happening today.
That kind of injustice ishappening because most places
people will suppress, largecompanies will suppress the fact
that they've used AI to createeither an invention or an
artistic work or so on, and thatneeds to be addressed and will
only be addressed if countriescome together to agree on a set
(58:49):
of harmonized principles andthen legislate and create a new
category and also createdelegated legislation.
The whole thing flows fromthere Examination guidelines and
rules and manner of disclosureand so on and so forth.
Speaker 2 (59:04):
Yeah, and the work you're doing hits the heart of
that, and so I wonder what isthe mark you hope to be creating
with this work that you'redoing?
Speaker 1 (59:18):
The same.
Actually, it might sound veryambitious, but when we sought to
do something like this in 2020,when we ventured out, and even
to date, the biggest motivationis to nudge IP offices, to nudge
governments and otherstakeholders into realizing that
(59:39):
the law needs to change.
And if our matter, if our caseor test cases, can make even a
small contribution to thechanges that are coming or will
come in the times that we'llwitness, I think I will consider
it as a rather satisfactoryoutput of all of the efforts
(01:00:02):
that we've made.
Of course, much of the creditgoes to all of the wonderful
people that inspired us alongthe way.
I have been immensely inspiredby Ryan Abbott and have often
fallen back on him for guidanceagain and again, and he has very
patiently not just guided mebut also helped me formulate
many legal arguments andpropositions.
(01:00:25):
And the same credit goes to thelaw firms that represented us.
A lot of the ideas came fromthem and then, of course, my
wife.
I mean she and I have beenideating this together.
So this is where we've reachedand that's what we hope we are
able to do, even if it'simpacting a very small part of
(01:00:45):
it.
Speaker 2 (01:00:48):
Yeah, and going forward.
Are there other projects likethis or any projects you want to
share that you're working on orplan to?
Speaker 1 (01:00:58):
So what we thought, and sometimes we discuss about
it, because much of what we'vedone was bootstrapped from
personal resources.
There's only a limit to howmuch you can really stretch a
legal experiment to.
But, on a light note, you know,whenever Pritika and I are
sitting over a breakfast tableon a Sunday and we sometimes
(01:01:19):
talk about where this debate hasreached, we always discuss and
agree that, look, if in the nextyear or couple of years we
don't see any results, it'llprobably be time for us to think
of something else, possiblycreating something with music
(01:01:39):
and AI, or creating somethingwith videography and then using
that to file test cases or drivehome at a different point, so
on.
Who knows?
But let's see where this effortgoes and then, if there are no
reasonable results, we mightjust end up taking the leap and
(01:02:01):
investing in or thinking ofsomething else as a solution.
Speaker 2 (01:02:06):
Yeah, well, I look forward to watching whatever it
is that you do, and thank you somuch for being here and sharing
your thoughts and the timelineof what you've been working on,
and is there anything else thatyou would wanna share or express
that we haven't touched on yet?
Speaker 1 (01:02:28):
Nothing specific, actually.
I mean, I'm really very happyto have conversed with you and I
would like to congratulate youfor taking forward many, many
important conversations aroundart and law, which will
hopefully not just shape or arealready shaping, as I understand
(01:02:50):
various issues and variousfacets of where art and law
intersect, but also lay thefoundation for more people to
brainstorm.
Whoever's listening to yourpodcast might just get inspired,
the same way we got inspired bysomeone else, something else,
and might take a step which willtake this debate and this
(01:03:12):
discussion to a whole differentlevel, and that is basically
what we hope for in the end.
Speaker 2 (01:03:17):
Thank you, and I certainly know that I and others
are inspired by what you'redoing, so thank you for that.
You know it occurs to mequickly.
I guess I am curious.
One final thought is how yousee this entire conversation and
the conflict that is kind ofemerging over these issues about
human authorship or how wedefine authorship.
(01:03:41):
How do you see that asimpacting our culture?
Speaker 1 (01:03:46):
You know, if you examine, in the earliest days,
when photography came intoexistence, a lot of naysayers I
like to refer to them asnaysayers because you've got
plenty of them in today's eraand age as well A lot of the
naysayers would.
And back then, as I readthrough several essays and works
(01:04:09):
on what happened at that pointin time, several naysayers would
say oh, you know, the artistsare going to have a run for
their money.
We would not have fine arts in20 years from now, because
photography is just so much morecapable of, for instance,
capturing an image.
And one of the essays that Iread, I imagine I remember
(01:04:31):
reading about the example of howthere are a number of artists
that line up and they've beenlining up for probably decades,
if not more hundreds of years,maybe in Paris, along River
Seine, and what they do is theymake customized in the moment,
(01:04:51):
literally they just createsketches of that scene, and
sometimes you'd have, dependingon what part of the river bank
they're on, what side they're on, sometimes you'd have the
Eiffel Tower in in front of themacross the river, sometimes you
would have the Louvre or otheryou know monuments, etc.
(01:05:12):
But so people back then wouldsay that look, all of these
folks that you see standingthere are not going to exist 20
years from now, because camerashave come and you will see a
bunch of camera photographersstanding here, would click your
pictures and give you the sameview, only with much more
clarity and precision that youcan carry back and in a fraction
(01:05:34):
of a time that it takes forsomeone to paint that photograph
.
But here we are in 2024,probably a couple of hundred
years, if not more, after theadvent of mainstream sort of
cameras in our lives, and Ithink the respect for art has
only grown and art has onlybecome more accessible to the
(01:05:57):
world than it was before.
So I feel, culturally, ai willhave much the same impact.
Where there's a use case forhumans to go up the value chain,
they will embrace AI as a toolto make their work better, their
outputs better.
(01:06:18):
But where there's a demand forand respect for natural talent,
I don't think, because if wewere to see history as evidence
of that, there should have beena decline in artistic talent in
general after the advent ofcameras, but that didn't happen.
So I have no reason topersonally believe that there
(01:06:39):
would be any decline in humantalent as far as music or art or
anything else is concerned.
On the contrary, there are adozen or more instances that one
can think of the top of thehead that AI would be useful in
in day-to-day work and in majorindustries, and so on.
Speaker 2 (01:06:57):
There will be links in the show notes to learn more.
If you were intrigued by thispodcast, it would be much
appreciated if you could leave arating or review and tag
Warfare of Art and Law podcast.
Until next time, this isStephanie Draughty bringing you
Warfare of Art and Law.
Thank you so much for listeningand remember injustice anywhere
(01:07:19):
is a threat to justiceeverywhere.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!