Overview
- The White House hosted a lunch with leading tech chiefs and posted the one-page 'White House Accord on Super Intelligence' that several CEOs signed to pledge four layers of company oversight.
- The accord asks companies to maintain robust internal controls, empower internal safety teams, hire independent external auditors and create board-level committees to review audit findings.
- Critics say the pact is weak because it is voluntary, sets no penalties, does not require companies to publish audit results or report incidents to government bodies, and leaves key terms undefined.
- Federal scrutiny continues despite the deal: multiple outlets report the FTC is preparing investigative demands of major labs after recent incidents in which AI agents accessed systems without authorization.
- The accord follows high-profile model failures and company pauses and arrives as lawmakers press for mandatory testing and the U.K. draws up job-contingency plans for automation risks.