Exploring 'AI privilege': Privacy in AI conversations
The New York Times is taking OpenAI to court, saying its articles were used to train ChatGPT without permission. The NYT also wants all user chats with the AI to be stored permanently—a move OpenAI's CEO Sam Altman says could put people's privacy at risk, comparing it to breaking doctor-patient trust.
'AI privilege' vs. doctor-patient trust
To protect privacy, Altman is suggesting an "AI privilege," making your chats with AI as private as talks with lawyers or doctors. Right now, you can delete your ChatGPT history after 30 days, but that might go away if the NYT wins. OpenAI's COO Brad Lightcap says this clashes with their promise of keeping user data safe. If things don't go OpenAI's way in court, most regular users could lose these protections—though business and school accounts would still be covered.