Six AI giants brought in auditors, but wrote their own exam.
On September 29, Nvidia, OpenAI, Anthropic, Google, Meta, and xAI pledged to the White House to accept external audits. But the agreement doesn’t say who picks the auditors, what they can examine, or whether the results will be made public—the companies decide what the auditors get to see.
How much are the safety promises of the AI you use every day worth?
Trust companies to regulate themselves, or wait for laws with teeth?
On September 29, Nvidia, OpenAI, Anthropic, Google, Meta, and xAI pledged to the White House to accept external audits. But the agreement doesn’t say who picks the auditors, what they can examine, or whether the results will be made public—the companies decide what the auditors get to see.
How much are the safety promises of the AI you use every day worth?
Trust companies to regulate themselves, or wait for laws with teeth?