AIQuiet · 37h
OpenAI Opens AI Models to Third-Party Safety Review Earlier in Development
The company will let outside groups evaluate its systems during training, evaluation, and deployment phases.
What to know
- OpenAI will enable third-party groups to evaluate its AI models for safety during training, evaluation, and deployment phases.
- The program represents an earlier intervention point than previous safety review practices.
- The move addresses growing concerns about potential harms from AI technology.
“OpenAI plans to let third-party groups vet its artificial intelligence models for safety risks in earlier phases of the development cycle, part of an ongoing effort to address heightened concerns about the technology's potential harms.”
OpenAI · Bloomberg Tech ↗
OpenAI AI company
How it unfolded 1 development · click the chart to see its coverage articlesposts
4 PMyesterday8 AM4 PMnow · 2:56 AM ET
-
1
-
first by Bloomberg Tech, 1d ago · also Bloomberg
-