Anthropic, OpenAI's proposed AI risk evaluators may not have enough power to prevent disasters
Source:
cnbc
September 16, 2026 · 07:31
From the publisher
Anthropic and OpenAI are proposing embedded AI evaluators to help manage risk of models causing catastrophic harm to society, but the idea has some issues.
The original article opens on the publisher's website.
More from Automotive
View topic →NASA Visits Schools Strengthening Florida’s Skilled Workforce
nasaNews
Sep 16, 2026 · 08:14
‘Godfather of AI’ says tech regulation is nearing Covid-style pivot moment
guardianTech
Sep 16, 2026 · 08:00
Meet a mouse whose brain cortex is made up of human cells
mitTechReview
Sep 16, 2026 · 08:00
Threads’ new features let podcasters promote shows and reach listeners
techcrunch
Sep 16, 2026 · 07:24
Small but mighty! Remembering the GameCube, Nintendo’s purple bundle of joy
guardianTech
Sep 16, 2026 · 07:00