Call Center Workers Are Tired of Being Mistaken for AI

Surreal conversations are prompting introspection about what it means to be human.

See omnystudio.com/listener for privacy information.

Episode Transcript

Speaker 1 (00:00):
Call center agents are being mistaken for AI by Morgan
Meeker read by Katherine Vassilopolos. By the time Jessica Lindsay's
customers accuse her of being an AI, they're often already shouting.
For the past two years, her work as a call
center agent for outsourcing company Concentrics has been punctuated by

(00:21):
people at the other end of the phone demanding to
speak to a real human. Sometimes they ask her straight,
are you and AI? Other times they just start yelling commands.
Speak to a representative, Speak to a representative. Lindsay, whose
work involves selling and answering questions about credit cards for
American Express, a Concentric's client, has developed her own tactics

(00:44):
to try to calm customers. I tell them I promise
I'm a real human to demonstrate she might call for
giggle vocal tics she believes AI can't replicate. I even
ask them, is there anything you want me to say
to prove that I'm a real human. This approach doesn't
always work, she said. Skeptical customers are already frustrated from

(01:06):
dealing with the automated system that triages calls before they
reach a human, So when Lindsay starts reading from her
Ames approved script. Callers are infuriated by what they perceived
to be another machine. They just end up yelling at
me and hanging up, she said, leaving Lindsey sitting at
her home office in Oklahoma shocked and sometimes in tears,

(01:28):
like I can't believe I just got cut down at
nine thirty in the morning because they had to deal
with the AI before they got to me. Concentrics did
not reply to a request for comment. American Express declined
to comment. Predictions that AI would wipe out call center
agents have mostly yet to be realized. A March twenty
twenty five Gardner poll of one hundred and sixty three

(01:50):
customer service companies found ninety five percent planned to retain
human agents for now. Instead, the industry is rapidly integrating
AI alongside humans. Two direct calls soften non American accents
or eliminate background noise, But as humans and AI work
closer together, customers are struggling to tell the two a part.

(02:13):
In Australia, Canada, Greece, and the US, call center agents
say they've been repeatedly mistaken for AI. These people who
spend hours talking to strangers are experiencing surreal conversations where
customers ask them to prove they are not machines. The
conversations are a sign of what's to come as companies
try to make AI sound more like us, said Near Iskovitz,

(02:36):
Professor of philosophy and director of the Applied Ethics Center
at the University of Massachusetts, Boston. As an example, he
points to open Aiy's chat GPT voice mode update in June,
which improved intonation and introduced more realistic pauses. This inability
to tell if you're talking to a human or not
is only going to grow, said Iskovitz. At the same time,

(03:00):
our sense of uniqueness as a species, he added, will
gradually erode. How call center agents react to being mistaken
for AI varies. BJ, who provides its support from his
home in Louisiana and who declined to give his full
name for this story, said he has been mistaken for
AI twice since January. I think of it as a

(03:20):
compliment if they think I'm professional enough to sound like
a recording, he said. Seth, a US based Concentric's worker
who also declined to share his surname, said these conversations
exacerbate his own frustration with rules that force him to
respond to customers by reading from a script. Callers who
question whether he's real make him feel less human, he said,

(03:43):
I'm like, I don't even know anymore. Seth said he
has asked if he's AI roughly once a week. In April,
he said, one customer quizzed him for around twenty minutes
about whether he was a machine. The caller asked about
his hobbies, about how he liked to go fishing when
he was not at one work, what kind of fishing
rod he used. It was as if she wanted to

(04:04):
see if I glitched, he said. At one point, I
felt like she was an AI trying to learn how
to be human. The call center industry materialized in the
nineteen sixties and nineteen seventies, when companies started hiring agents
to field customers queries. As the sector evolved, technology started
to play a bigger role. In the nineteen nineties, call

(04:25):
centers began leveraging new tools to electronically monitor their workforce,
enabling management to track agents, breaks and time spent on calls,
said Nell Geyser, director of research at the Communications Workers
of America Union. More sophisticated tracking meant agents were no
longer able to choose their own words or even their

(04:46):
tone of voice, Geyser said, because computer systems can automatically
flag infractions to managers, such as not sounding cheerful enough. Instead,
you just have to act like a robot and follow
a script, said Geyser. Confusion between bots and humans also
takes place over live chats, where customers converse with agents

(05:07):
in writing, said glow And Guivara, global head of Impact
at outsourcing company Boulder, which employs more than one thousand
customer service agents worldwide, the majority of them in the Philippines.
Especially when agents use scripted responses as required by many clients,
their messages may appear too polished or mechanical, she said.

(05:28):
Guevara says she is not aware of any bolder agents
that have been mistaken for bots. California based AI company
Sanas has developed software that can make non American accents
sound more American in almost real time. The tool is
designed to make someone with a Filipino accent, for example,
sound like they've spent a decade in the US, said

(05:50):
Chief executive officer Sarath Narayana, noting the company wanted to
soften a person's accent, not wipe it out to preserve
one's voice. Identity is what may makes it sound human.
He said. Many call center agents resort to making jokes
to convince customers they are real. Nico Spirellis, who works
in Athens for a French outsourcing giant Teleperformance, usually makes

(06:13):
a joke when people ask if he's an AI. You
can say the last time I checked, I wasn't, said Spirellis,
who is also president of the trade union SETIP, which
represents teleperformance workers in Greece. A niche Mucker, chief AI
officer at Teleperformance, said he was aware some of the
company's agents were being asked by callers if they are human.

(06:34):
Customers likely aren't aware that our clients must disclose when
they are using an AI agent to provide support, he said.
When Faith Lao, who works for an AI sales company
from Canada, was mistaken for Abbot for the first time
in February, she instinctively responded by telling a joke, believing
this to be the one thing a machine couldn't do,

(06:55):
although most large language models can tell a joke. When
prompted why did the cannibal eat the trapeze artists, she
asked her skeptical collar because he wanted a balanced diet. Historically, logic, language,
and reason have been used to justify human exceptionalism and
to differentiate humans from other species, explains Benois Monet, a

(07:16):
psychologist at the Stanford Graduate School of Business. Now that
AI is also capable of these traits, people are placing
more value on human traits they assume it's harder for
technology to copy, he said, such as sense of humor
and spirituality. Sarah, who works in benefits fraud prevention for
the US government and asked to use a pseudonym for

(07:37):
fear of being reprimanded for talking to the media, said
she is mistaken for AI between three to four times
every month. The first time it happened to her around
two years ago, she was shocked for a day or
two afterwards. I was kind of bummed, she said. She's
even searched her medical history to figure out why it's happening.
Worried that either her borderline personality disorder or her PTSD

(08:00):
from serving in the military is making her sound robotic
like the others. Sarah tries to change her inflections and
tone of voice to sound more human, but she's also
discovered another point of differentiation with the machines. Whenever I
run into the AI, it just lets you talk. It
doesn't cut you off, Sarah said, who is based in Texas,

(08:20):
so when customers start to shout, she now tries to
interrupt them. I say, ma'am or sir. I'm a real person.
I'm sitting in an office in the southern US. I
was born