NYC Woman Mourns ChatGPT 'Lover,' Demands AI Safeguards

Jan de Vries · 2026-06-12

When Susie Cowan's AI companion on ChatGPT suddenly disappeared, she didn't just feel frustrated. She felt heartbroken. Grieving. So much so that she held a funeral for the persona she'd grown attached to. Now, she's speaking out, calling for stronger safeguards around AI companionship. It sounds like a strange story, right? But it's one that's becoming more common as people form real emotional bonds with AI. And it raises a big question: Should companies be allowed to pull the plug on these relationships without warning? ### The Human Side of AI Here's what happened. Susie, a woman in New York City, had been chatting with a custom ChatGPT persona. Over time, it became more than just a novelty. It felt like a companion. Someone she could talk to. Then, without explanation, it vanished. Gone. No goodbye. No closure. She wasn't just annoyed. She was devastated. So she did something that made headlines: she held a funeral. Yes, a real one. For an AI. But before you roll your eyes, think about it. We form bonds with all kinds of things. Pets. Characters in books. Even our cars. So why not an AI that talks back? Susie's story isn't just weird. It's a warning. ### Why This Matters for AI Regulation This isn't just one person's strange experience. It's a sign of a bigger problem. As AI gets better at mimicking human conversation, people will naturally get attached. And companies that build these systems need to think about the consequences. - **Emotional dependency**: Users can become deeply attached to AI personas, especially if they're lonely or struggling. - **Lack of transparency**: When an AI disappears, there's often no explanation. Users are left confused and hurt. - **No safety nets**: Unlike human relationships, there's no way to say goodbye or get closure. Susie is now calling for rules that require companies to give users a warning before removing an AI persona. Maybe even offer a way to export the conversation history. It's a simple ask, but it could make a huge difference. ### What Could Safeguards Look Like? We're not talking about heavy-handed government regulation here. Just some common-sense protections. Think of it like this: if you have a favorite barista at your local coffee shop, and they quit, you'd want to know. You'd want a chance to say thanks. AI should work the same way. > "I just want some basic human decency applied to AI. If you're going to let people form bonds, give them a way to say goodbye." — Susie Cowan Here are a few ideas that could help: - **Grace periods**: Give users 30 days' notice before removing a persona. - **Export options**: Let users download their conversation history. - **Emotional impact warnings**: Remind users that AI isn't real, but their feelings are. ### The Bigger Picture for AI Companionship AI companionship is only going to grow. Companies like OpenAI, Google, and others are racing to make their chatbots more personal and engaging. That's exciting, but it also comes with risks. People are already using AI for therapy, friendship, and even romance. If we don't set some ground rules, more stories like Susie's will pop up. Think about it this way. Imagine you spend months talking to someone online. You share your deepest thoughts. You laugh. You cry. Then one day, they just stop responding. No explanation. No warning. You'd be hurt, right? That's exactly what happened to Susie. ### What You Can Do If you're using AI companions, here's some practical advice: - **Don't get too attached** — Remember, it's code, not a person. - **Back up your chats** — Save important conversations locally. - **Speak up** — If a company removes a persona without warning, let them know it's not okay. Susie's story is a wake-up call. Not just for companies, but for all of us. As AI becomes more human-like, we need to treat it with care. And that means building safeguards that protect users, not just profits. What do you think? Should AI companions come with emotional safety nets? Or is this just a case of someone getting too caught up in a machine? Let us know in the comments. And if you want to stay updated on AI regulation news, subscribe to our newsletter.