Anthropic and OpenAI turn to third-party safety evaluators as funding rules stay unsettled
Overview
Anthropic and OpenAI are seeking support from small third-party safety evaluators, including METR, Apollo Research and Transluce, CNBC reports.
Federal regulation of AI remains absent, so the arrangement has no national framework.
CNBC says key questions are unanswered: how the evaluators should be funded, what level of access they will receive, and what their reporting structure will be. California has passed laws on independent verification organizations, and an OpenAI employee dismissal has raised concerns about third-party communication, according to the same report.
OpenAI disputes CNBC's characterization. In a post, it says it is "actively finalizing contracts with third-party safety assessors and will announce details in the coming weeks."
Written by AI from the articles below · updated Oct 11, 7:56 AM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- CNBC · TechnologyNewsIndependent AI evaluators gain a bigger role as funding and rules remain unsettled
AICNBC reports that Anthropic and OpenAI are turning to small third-party evaluators such as METR, Apollo Research and Transluce as federal regulation remains absent. Questions over how those evaluators are funded, what access they receive and how they report findings remain unresolved, while California has passed laws on independent verification organizations and an OpenAI employee dismissal has raised concerns about third-party communication.
Heat trend
Not enough continuous observations to show a trend yet.