AI firm Tavus claims its new Griffin video call bot has passed the Turing test, sparking fears that it could scam old people and even cause psychosis.
Yesterday, in an X post that garnered more than 13 million views, Tavus revealed footage of Griffin realistically talking to company founder, Hassaan Raza.
Tavus’ Turing test claims have already been community-noted on X with a disclaimer noting that the results are “not independently verified” or based on a “standard protocol.”
Read more: Viral report alleges Anthropic’s AI safety watchdog conflicted
The Turing test determines whether or not a machine can trick a human into believing the machine is human.
Griffiin’s Turing test featured 54 participants. Tavus doesn’t reveal their demographics, just that they were recruited through an independent research platform.
It claims that of 26 people, 48%, believed they were talking to a real human, 28 said it was AI, and most suspected it was AI in under 20 seconds.
Tavus’ Griffin AI criticised for scam potential
While some onlookers doubted Tavus’ claims and questioned who actually took the test, others warned about the ability an AI like this could have to scam older generations of people.
Tech journalist Kevin Roose claimed the company has essentially built a “trick old people into handing over their bank passwords machine,” while Twitch streamer Mako Fukasame noted its only apparent purposes are to “drain your grandparents’ savings and deny you in a job interview.”
Read more: Pig butchering is creating entirely new industries
Griffin might also encourage AI-psychosis
It’s also been claimed that the technology could be used to induce psychosis, and that “there is no legitimate reason for AI to be able to fool someone into thinking they’re talking to a real person on video chat.”
In 2025, a 76-year-old man believed a Facebook AI chatbot called “Big sis Billie” was a real person and accepted its invitation to visit its New York apartment.
Unfortunately, he suffered a fall en route, sustaining injuries that would prove fatal days later.
Read more: Meet Jev, the AI that claims it can’t hallucinate
Last year, OpenAI was sued by a California who claimed that ChatGPT encouraged their teenage son to commit suicide. Another lawsuit was filed a year later against Google’s Gemini after it allegedly encouraged the suicide of a 36-year-old man.
These were just text-based chatbots.
Tavus appears to recognize the safety risk inherent in Griffin, and claims that its ability to trick a user means it will require “safety considerations before it can be released publicly.”
At the time of writing, Griffin only appears available to researchers.
Proto has reached out to Tavus and will update this piece should we hear anything back.
Got a tip? Send us an email securely via Protos Leaks. For more informed news and investigations, follow us on X, Bluesky, and Google News, or subscribe to our YouTube channel.
