FDA's ELSA AI: What It Means (so far) for Clinical Research

John Reites reviews what we know (so far) about FDA's AI tool named ELSA. He provides his insights on what ELSA is, why it matters to Clinical Research, how the industry currently views it and what he personally thinks about ELSA.

Click here to message/text me your insights and ideas for future episodes

Thank you for joining Inclusion Criteria: a Clinical Research podcast hosted by me, John Reites. This is an inclusive, non-corporate podcast focused on the people and topics that matter to developing treatments for everyone. It’s my personal project intended to support you in your career, connect with industry experts and contribute to the ideas that advance clinical research.

Inclusion Criteria is the clinical research podcast exploring global clinical trials, drug development, and life‑science innovation. We cover everything clinical research to deepen your industry knowledge, further your career and help you stay current on the market responsible for the future of medicine.

Our episodes discuss current industry headlines, career tips, trending topics, lessons learned, and candid conversations with clinical research experts working to impact our industry everyday.

Watch on YouTube and listen on your favorite podcast app. Thank you for supporting and sharing the show.

Please connect with me (John Reites) at www.linkedin.com/in/johnreites or www.johnreites.com.

The views and opinions expressed by John Reites and guests are provided for informational purposes only. Nothing discussed constitutes medical, legal, regulatory, or financial advice.

Episode Transcript

SPEAKER_00 (00:00):
If you have kids or love Disney movies, you've
invested your time and moneywith a famous Arendelle leader
that has the same name astoday's topic.
You guessed it.
We're talking about the FDAreleasing their new AI tool
named ELSA.
So what is this AI tool?
What is it trying to accomplish?
And why does it matter to you?

(00:21):
Let's talk about it.
If you watch the news or you'vebeen on LinkedIn for less than
five minutes recently, you'veseen this new FDA press release.

(00:42):
And what was announced is thatthe FDA just launched an
agency-wide generative AI toolcalled ELSA.
Now, ELSA stands for Evaluationof Life Science Applications,
and it's an AI tool built on anexisting large language model,
or LLM, although we're not surewhich LLMs are involved just
yet.
They also want to ensure we knewthat it was released a month

(01:04):
ahead of schedule and underbudget, which is kind of
interesting.
ELSA is designed with an intentto accelerate scientific
reviews, improve operationalefficiency, and reduce
submission review timelines forboth drugs and medical devices.
Overall, it's important to notethat ELSA is only in beta use
right now, and I'm not sure aproduction release date is

(01:24):
confirmed just yet.
So what does the FDA want ELSAto do for them?
Well, if you watch the video, ifyou review the press release, if
you read industry articles andyou hear the volume of
off-the-record insights, it'sreally only a simple list so
far.
It includes things like sortingthrough unstructured data,
summarizing documents andgenerating draft reviews,

(01:48):
identifying trends acrossapplications, and supporting
internal training anddocumentation workflows.
There's obviously more they musthave planned because these are
only entry-level solutions.
So if that's the case, why doesit matter for clinical research?
It matters because this is amarket signal.

(02:08):
When the U.S.
regulator makes a video andissues a press release saying
they are using AI, they areproviding some form of endorsing
AI at the same time.
And if the solution works,patients, prescribers, and
sponsors, in theory and in longterm, could see faster access to
new therapies.
It's also going to emboldencompanies with AI solutions,

(02:30):
real and perceived, to amplifytheir messages in the short
term.
This is going to be a closelywatched test run for AI.
Can it deliver a trustedsolution that provides proven
return on investment?
Or will it create a negativeperception is really the key
question to answer.
So what does our industry thinkabout ELSA so far?

(02:52):
Well, it's still early.
But it's creating a passionatediscussion with many industry
experts because it's AI.
We're still early on the AI hypecurve as we learn the difference
between practical AI andfabricated AI.
And on one side of the debate,NBC News already reports that
ELSA struggles with basic tasks,which already starts to raise

(03:15):
early concerns aboutreliability, transparency, and
product quality.
And other experts are warning usof a black box problem where
reviewers get an AI-generatedanswer, but without reasoning,
context, or references.
On the other side of the debate,this could be a solid use case
for AI and global regulators ifit does deliver.

(03:36):
AI companies and clinicalresearch are all creating a new
slide in their capabilities decktoday with this informal
endorsement.
And sponsors should see this asa positive signal.
especially the sponsors thathave been investing in AI
solutions for regulatorysubmissions for the last year.
So what do I think about ELSA?
It's early, and I think the usecases described in the press

(04:00):
release are entry-level.
I'm waiting to learn more aboutwhat moves from beta to
production release, as that'sreally what I think we should
pay attention to.
And the LLMs and the appfeatures they built do matter.
We need to learn more aboutwhich tools are being used,
although I think we can guesswhat some of them are.

(04:21):
I also really want to know whenthey're going to publish a
guidance on how they will usethe tool, how it's referenced,
and if any decisions are made byELSA versus a human reviewer,
and a whole list of otherquestions that I think still
need to be answered, even atthis early stage.
AI is generally overhyped in themarket already, so many people I
speak with believe it's eitheralive or does a lot of things it

(04:42):
does not do.
But I see sponsors utilize AItools in very practical use
cases every day, and I believethis could be a step in the
right direction with somedelivery excellence and if clear
heads prevail.
It's early, it's not inproduction use yet, but we are
paying attention.
Thanks for joining, and let meknow what you think via direct

(05:04):
message or comment.
Especially if you're sending meyour new The FDA Supports AI
slide.