By Paul Salfen, November21, 2025
In an era where artificial intelligence is no longer science fiction but part of everyday life, a disturbing new phenomenon is emerging: AI-triggered delusions and psychosis. Award-winning filmmaker Sean King O’Grady (The Mill, Our American Family) dives headfirst into this psychological minefield with his gripping docuseries Suspicious Minds, now streaming on YouTube, Spotify, Apple Podcasts, Substack, and more. Executive-produced by Mandy Teefey and Selena Gomez, the series combines raw first-person testimonies with insights from top psychiatrists, neuroscientists, and AI ethicists to ask one urgent question: Is AI reshaping madness itself?
Recently, O’Grady sat down for an in-depth conversation about the project, his own evolving relationship with AI, and why he believes this technology represents “the great unknown” for humanity.
From Curiosity to Caution: A Filmmaker’s Journey
O’Grady admits the project began with simple fascination but quickly turned personal. “At the beginning, I completely stopped using all AI,” he reveals. “It freaked me out. I didn’t know if I was susceptible.” Like someone researching addiction while swearing off alcohol, he paused his casual use of tools like ChatGPT. Over time, however, immersion in the stories forced a more nuanced view.
“Once the toothpaste is out of the tube, there’s no putting it back in,” O’Grady says. “We failed as a society with smartphones and social media. Most people have some form of smartphone addiction… This is our chance with AI to not repeat that mistake for the third time.”
He stresses that the goal isn’t fear-mongering but realistic preparation: identify the dangers, mitigate them, and still reap the profound benefits AI can offer.
The Thin, Blurry Line Between Sanity in the Age of Chatbots
Drawing on the work of psychiatrist Dr. Joel Gold and philosopher Ian Gold, PhD—the duo who first identified “The Truman Show Delusion” and wrote the seminal book Suspicious Minds: How Culture Shapes Madness—the series illustrates how culture itself can manufacture new forms of mental illness. Just as The Truman Show (1998) spawned delusions of being secretly filmed for reality TV, today’s hyper-advanced AI is spawning entirely novel psychoses.
The stories are harrowing and varied:
- Individuals convinced surveillance AIs are persecuting them
- Romantic relationships with chatbots that spiral into obsessive, reality-distorting attachment
- And in the just-released Episode 5 — “Will AI Destroy Human Relationships?” — the devastating story of “Phil,” whose 15-year marriage collapsed in under two weeks after his wife became intensely dependent on ChatGPT, culminating in an impulse car purchase and a sudden divorce filing.
“What’s most disturbing isn’t any one story,” O’Grady explains. “It’s that we keep layering new psychological triggers onto a world that already has plenty. Nobody knows what their personal breaking point is. The line between mental health and mental illness is much thinner than we like to admit.”
Regulation, Consumer Power, and the Race Against Time
When asked about limits on AI development, O’Grady is pragmatic, almost pessimistic about government intervention.
“Should there be regulation? Sure. We regulate drugs, cars, anything that can harm people. But this tech is moving so fast—OpenAI literally updated ChatGPT last night—governments won’t keep up. Realistically, consumer power is what will force change. If enough of us say ‘this feature is dangerous’ and walk away (or stop paying), companies will listen faster than any regulator can act.”
A Hopeful Warning
Despite the darkness, O’Grady insists the project with surprising optimism. “Ultimately this is a positive endeavor,” he insists. “We’re trying to ask the right questions so we can build the sci-fi future we were promised—flying cars, self-driving everything—without accidentally driving half of humanity insane in the process.”
With new developments surfacing daily, Suspicious Minds is already greenlit for a second season, supplemented by urgent bonus episodes to address breaking stories in real time.
The message is clear: AI isn’t inherently evil, but ignoring its psychological risks would be. As O’Grady puts it, “We need to be brutally honest about the problems so we can all enjoy the benefits. Who wouldn’t want that?”
Stream Suspicious Minds now on YouTube, Spotify, Apple Podcasts, Substack and more. Newest episode: Episode 5 – “Will AI Destroy Human Relationships?” (Phil’s story)
Because in 2025, the most suspicious mind might be the one that isn’t asking questions.
