Six AI companies pledge to accept voluntary White House safety audits
Start here
AI-assisted source synthesis · not an independent investigation.
On September 30, 2026, OpenAI, Google, Meta, Anthropic, Nvidia, and xAI agreed, under a voluntary White House agreement, to bring in external auditors to assess AI safety controls. The agreement calls for monitoring whether advanced models could facilitate cyberattacks, hacking, and biological or chemical threats, and for checking whether related safeguards are effective. However, the agreement has no enforcement mechanism, penalties, disclosure requirements, or implementation deadlines; companies choose the auditors themselves and are responsible for remediation.
AI-written from reports · updated Sep 30, 2026
Timeline
Follow the reporting to see the different sides of the story.
- CoinDeskOpenAI, Google, Meta and others commit to accepting independent AI safety audits under a voluntary White House agreement
OpenAI, Google, Meta, Anthropic, Nvidia and xAI have agreed to bring in external auditing firms to assess AI safety controls under a voluntary White House agreement. The agreement requires monitoring whether advanced models could enable cyberattacks, hacking, or biological or chemical threats, and checking whether safeguards are effective. The agreement includes no enforcement mechanisms, penalties, disclosure requirements or implementation deadlines; auditing firms will be selected by the companies themselves, and the companies will be responsible for remediation.