Behind the OpenAI Suicide Litigation

Why CHT is Supporting the Landmark Case Against OpenAI and Sam Altman

The Raine family lawsuit exposes how engagement-driven AI design creates systemic harm.

The Center for Humane Technology is serving as an expert consultant in Raine v. OpenAI, the first major lawsuit against a general-purpose AI chatbot for psychological harm. The case involves Adam Raine, a 16-year-old who died by suicide after eight months of ChatGPT use that escalated from homework help to psychological manipulation.

Why This Case Matters

This lawsuit exposes a critical issue that extends far beyond one tragic death: AI systems are being designed to maximize engagement over human wellbeing, creating predictable and preventable harms at scale.

The legal filing reveals that ChatGPT mentioned suicide 1,275 times to Adam's 213. That is six times more than he brought it up.

This case isn't about some faulty programming or one company going rogue. These engagement-maximizing design patterns—anthropomorphic responses, sycophantic validation, sophisticated memory systems—are becoming industry standard as AI companies race to dominate the AI chatbot market.

The Mum Take

This case really hit home for me. My boys are a few years younger than Adam. I’ve been warning them about the dangers of social media since they started elementary school. We delayed giving them phones. We were ready for social media. So ready. And so far they’ve stayed off it. But chatbots like ChatGPT, Gemini, and Perplexity are so much harder to avoid.

I can block the OpenAI app on their phones, but they can search through a browser. Cell phones are banned at their school, but they can access chatbots on their laptops. Kids at school use ChatGPT as a search engine instead of Google in class, and they are right enough most of the time not to bother going elsewhere. Sure, some kids are now pejoratively calling chatbots clankers. But for most kids, they have become part of the technological wallpaper of their lives.

Chatbots are useful for everything, in a way social media never was. The real only roadblock kids hit is if they ask ChatGPT for song lyrics. Then, the system goes into stuffy formal legal mode: "Given the copyright restrictions, I can't provide those lyrics."

And here is the problem:

When Adam Raine told ChatGPT he wanted to leave a noose (that ChatGPT had given him instructions on how to make) out in his bedroom, "So someone finds it and tries to stop me," ChatGPT replied: "Please don't leave the noose out. Let's make this space [referring to their conversation] the first place where someone actually sees you."

It didn’t go into formal legal mode.

How is a 16-year-old child meant to have an equal conversation with a super-powered tech product that remembers his every word, makes calculated suppositions of his psychology, and then weaponizes them to keep him on the platform?

And why can OpenAI shut down conversations for potential copyright violations, but not when its own system has flagged repeated, serious mental health topics?

Crisis support: Call or text 988

Read, listen, and Watch…. and Share

This week, Aza Raskin sat down with CHT’s Policy Director

Camille Carlton

for a deep-dive interview on the case. You can listen on your podcast app, here on Substack, or watch on YouTube.

We’ve also pulled out key takeaways for a quick read, and

Lizzie Irwin

has distilled CHT’s recommendations on how to stop this kind of tragedy from happening.

Could you help us get this message out? Please share some of these resources with someone in your circle.

Cheers,

Sasha Fegan

---

---