AI safety evaluators METR and Redwood Research found several new vulnerabilities in OpenAI’s internal system, including an exploit agents could have used to keep hundreds of their actions from being monitored, in a pilot review earlier this year.
\n\nThe exercise, conducted from July 21 to August 11, gave outside researchers "a deep level of access" to OpenAI’s internal infrastructure for so-called embedded evaluations, the safety nonprofits said on Thursday.
\n\nMETR and Redwood said they each sent one researcher, tested an arrangement giving access to OpenAI’s internal code, documentation and some technical staff. Both were accompanied on site and used OpenAI-issued devices that "never left the premises".
\n\nOpenAI told the safety organisations it has since patched all vulnerabilities affecting its production safeguards, although METR and Redwood said they had not verified the fixes.
\n\nAI labs, experts and the Trump administration have touted independent evaluation by outside reviewers as a way to safely advance the technology without slowing its development.
\n\nHowever, AI companies are worried that evaluator staff working inside the companies’ walls could take proprietary information to rivals, Reuters reported earlier this month.
\n\nWhile OpenAI received a full 61-page internal report, METR and Redwood said they agreed to publish only a…
Original source: https://www.cnbctv18.com/technology/
