OpenAI Dissolves Catastrophic AI Risk Team
TL;DR. OpenAI disbanded its 'Preparedness' team responsible for evaluating catastrophic AI risks, reassigning its work to existing groups. - The team focused on biological and cyber risks posed by OpenAI's own models. - Former team lead now focuses on recursively self-improving AI safety risks. - Several safety staffers have departed, leading to internal unease about safety commitment.
- OpenAI dissolved its 'Preparedness' team, which assessed catastrophic AI risks.
- The team's work, including biological and cyber risks, is now integrated into other groups.
- Several safety personnel, including the Chief Ethics Officer, have recently left OpenAI.
- Internal sources report growing concern over OpenAI's commitment to safety measures.
Sources
- OpenAI dissolved the team built to catch catastrophic AI risks, reassigning its work to other groups — the-decoder.com
- theverge.com — theverge.com
- analyticsindiamag.com — analyticsindiamag.com
- engadget.com — engadget.com