Necessary AI privacy guardrails could only be determined by ensuring a dynamic response to emerging problems, said Altman at an International Association of Privacy Professionals conference, where he also referred to the privacy protections as a responsibility of society rather than AI firms.
"We don't have [privilege] yet for AI systems, and yet people are using it in a similar way, and that is a place where I think society will have to come up with a new sort of framework. The way that I believe works best for that is a very tight feedback loop and sort of watching what's happening in people's lives and iterating in response to that," Altman added.




