Is GenAI Addictive? A Conversation with James Bedford

Why are we so afraid to use this word?

Overreliance. Overdependence. Compulsion. All-consuming obsessions. AI-induced psychosis.

All of these terms are used freely when discussing the risks and dangers of GenAI use. For some reason, the conversation rarely veers into the language of addiction.

I think that is a mistake.

GenAI appears to have highly addictive qualities. I am not a psychologist, psychiatrist, counselor, or trained mental health expert - but I certainly feel like I might be getting addicted. Areas where I used to feel confident in my own skills and abilities — like writing concise, thorough, balanced emails — have now become areas where I consistently reach out to AI for feedback. The 2015 version of me would be quite disturbed.

Is this what researchers are calling “GAID”? The thought that I might be developing “Generative AI Addiction Syndrome” scares me more than I care to admit.

That's how I ended up chatting with James Bedford, an AI Education Specialist at the University of New South Wales, PhD in Creative Writing, and novelist. His "No AI December" project caught my attention immediately.

Adherents to “Dry January” in the States (or “Dry July” down under) will recognize the nomenclature. James’ idea was to abstain from AI use for an entire month - whilst keeping a journal of his thoughts and changing work patterns. What came from it were fascinating reflections and a plan to crowdsource the initiative in December 2025 for a research project.

Academic researchers are starting to take notice. In March, the Asian Journal of Psychiatry published a Letter to the Editor titled “Generative artificial intelligence addiction syndrome: A new behavioral disorder?” In it, they write, “Unlike conventional digital addictions, which are often characterized by passive consumption, GAID involves compulsive, co-creative engagement with AI systems, leading to cognitive, emotional, and social impairments. A fundamental distinction must be drawn between dependence and addiction.”

This distinction became central to my conversation with James on the inaugural episode of The Educator's Dilemma podcast. When does dependence become addiction? And why are we culturally refusing to recognize and talk about the possibility that we may be dealing with an entirely new form of addiction?

Here’s the video - let me know your thoughts in the comments on YouTube or here on Substack. And don’t forget to like, subscribe, and share with friends and colleagues.

For more on this subject…

1. Lance Eliot wrote about this over a year ago, though I can’t remember it creating more than a ripple.

2. The Guardian argues that we are facing a new set of “cognitohazards.” Might need to work on that name, but hey — shoot your shot, Guardian.

3. Vox’s article published only a week ago shines an important light on the hazards of compulsion and addiction for those already diagnosed with OCD.

Still, it's very early days on this subject, which is exactly why James' research project feels so important. I'll be participating in No AI December 2025, documenting my own journey of abstinence and reflection (even if it means I have to use AI to search for Christmas gifts in November.) I hope you'll consider joining us - imagine the collective insights we could gather if we all stepped away from our AI tools together for a month and shared what we discover about ourselves, our work, and our relationships with these technologies. The vulnerability might be uncomfortable, but the learning could be transformative.