Trump may be forced to reveal secret rules feds use for AI safety testing
Trump’s secret reviews of frontier AI models may hide corruption, lawsuit says.
TL;DR
- Protect Democracy has sued four federal agencies over the Trump administration's secret AI safety review framework.
- The lawsuit claims that almost no details about the framework, including participating companies and legal authority, have been released to the public or Congress.
- Concerns exist that the framework could be used corruptly, potentially favoring AI firms favored by Trump and retaliating against others.
- The group is seeking an order for the release of unclassified information about the framework, including its text, terms of participation, and criteria for granting access to AI models.
- Lawmakers, including US Representative Greg Casar and California state Senator Josh Becker, have also voiced concerns about the lack of transparency and oversight.
- The White House launched the GOLD EAGLE clearinghouse, which also lacks publicly identified participants or stated legal authority.
- Protect Democracy argues that Congress cannot meaningfully oversee programs whose terms and authority are not disclosed, impacting decisions on renewing legislation like CISA.
- The organization fears the framework could be used to coerce AI companies, extort support, or force politicized viewpoints into AI models.