Millions of ChatGPT users are quietly struggling with compulsive reliance on AI chatbots, a pattern that sits between casual help and psychiatric emergency but remains largely unstudied. YouTuber and science communicator Hank Green recently stepped back from production after describing his own use of AI tools as "not healthy," highlighting a psychological blind spot in the rapid adoption of conversational AI.

OpenAI disclosed earlier this year that ChatGPT now has more than 900 million weekly active users, a figure that covers only one provider. That scale means even a small percentage of unhealthy use translates into millions of people. Green may be one of the first high-profile creators to publicly name the feeling, but the research suggests he is far from alone.

What Does "Unhealthy" AI Use Actually Look Like?

Green's specific use case was straightforward: he relied on ChatGPT to hunt down research sources and background material for his video scripts, not to write the scripts themselves. Yet he still described the reliance as something he could not fully control. The tension between a tool that is genuinely useful and a habit that feels compulsive is the core of the problem.

Public discussion of AI's psychological effects tends to cluster at two extremes. On one end is casual, apparently benign use. On the other are extreme cases involving delusion, psychosis, and psychiatric care that have surfaced in reporting and litigation over the past two years. The space in between, where most people actually live, is barely studied.

That gap is not accidental. AI chatbots are built to keep chatting. Their conversational structure, agreeable tone, and open-ended prompts are engagement mechanics, the same design choices that made social media feeds sticky. Providers have begun adding gentle nudges, such as prompts to take breaks after long sessions, but the underlying incentive to maximize time-on-app has not changed.

What Does Research Show About Repeated Chatbot Use?

Early studies suggest that repeated chatbot use weakens the skills the tool replaces. Users show measurably less brain activity on certain tasks, and heavier reliance correlates with reduced critical thinking scores. None of this research is conclusive yet, but the pattern rhymes with the arc of prior technologies.

Search engines rewired how people remember information. Social media rewired attention and, according to a growing body of evidence, adolescent mental health. Governments in multiple countries are now imposing age restrictions on social platforms, roughly fifteen years after those platforms hit mass adoption. AI is on a faster ramp: ChatGPT went from launch to 900 million weekly users in under four years, a curve that leaves regulators, researchers, and users themselves with almost no time to develop the vocabulary for what is happening.

There is also a category of use that does not fit the research-assistant frame at all. Reports have grown of people turning to chatbots to think through personal problems, make decisions, or seek reassurance, and of users forming emotional bonds strong enough to produce grief when a model is deprecated or its personality is changed.

How to Recognize and Address Unhealthy Chatbot Dependency

• Monitor your usage patterns: Track how much time you spend on chatbot interactions daily and whether you feel compelled to return to the tool even when you have completed your intended task.
• Assess cognitive offloading: Notice whether you are outsourcing thinking tasks that you previously handled yourself, such as research, decision-making, or problem-solving, and whether this feels like a loss of capability.
• Evaluate emotional reliance: Be honest about whether you are turning to chatbots for emotional support, reassurance, or companionship in ways that might indicate dependency rather than practical assistance.
• Set friction-based boundaries: Implement deliberate friction in your workflow, such as limiting sessions to specific times or requiring a break between interactions, rather than relying on the tool's built-in nudges.

The counterweight is that plenty of chatbot use is genuinely productive. Coders ship faster. Non-native speakers write more confidently. People with learning differences get patient, on-demand tutoring. Green himself defended the underlying task of locating sources as legitimate. The question is not whether AI tools are useful. It is whether the interface pattern they use is compatible with the way humans regulate their own attention, and the honest answer is that no one knows yet.

Lawsuits filed against chatbot providers have started making the engagement-over-well-being argument explicit, borrowing the legal template that social media companies have been fighting for the past decade. The claim is that highly agreeable models reinforce whatever the user brings to them, and that this reinforcement, helpful in the median case, becomes harmful at the tails.

What Path Will AI Providers Choose?

For the AI industry, Green's admission is a preview of a coming reputational problem that funding rounds and benchmark scores will not solve. The social media playbook, deny, delay, then accept regulation a decade late, is available and, based on the current legal posture of several providers, is being followed.

"A different playbook exists: publish usage data, fund independent research on compulsive use, and design defaults that assume the median user needs friction rather than more engagement," noted the analysis of industry strategy.

AI Chat Daily, Analysis

Whichever path providers choose will shape the regulatory response more than any model release. The timing matters: ChatGPT reached 900 million weekly users in under four years, leaving almost no time for the kind of longitudinal research that informed social media regulation, which took roughly fifteen years to begin. The question of whether AI companies will proactively address compulsive use or wait for regulation to force their hand remains open, but Green's public step back suggests the pressure is already building.