Frontier risk and preparedness
OpenAI Blog ● Covered by 2 sources
Frontier AI is developing methods to prepare for catastrophic risks from advanced AI systems by establishing a dedicated preparedness team and launching a challenge initiative. No specific metrics, timelines, or budget allocations are disclosed in this announcement. The effort aims to create organizational structures and external engagement mechanisms focused on potential severe outcomes from frontier models.
Why it matters
To support the safety of highly-capable AI systems, we are developing our approach to catastrophic risk preparedness, including building a Preparedness team and launching a challenge.
