Frances Haugen Questions 7 AI Firms' White House Pledge as Self-Regulation Faces New Scrutiny

3 articles · Updated · CNBC · Oct 2

Frances Haugen said AI companies must comply with the spirit—not just the letter—of the White House self-regulation agreement signed days ago by Mark Zuckerberg and other CEOs.
On CNBC, Haugen warned flexible AI systems create room for companies to exploit gaps in loosely written rules and "go around the end of the fence."
She said AI firms have been more proactive on safety than past tech companies, but pointed to Anthropic CEO Dario Amodei's push for independent auditors as recognition that companies can still cut corners.
Haugen's comments draw on her 2021 Facebook leaks, which showed Meta knew of harms including negative effects on teens' mental health, underscoring her broader argument that voluntary oversight needs outside checks.

Sources

Left50%

Center50%

What hidden AI safety risks are tech giants currently masking behind the vague language of voluntary commitments?

Could mandatory third-party AI audits stifle innovation, or are they the only way to prevent another tech crisis?

How will regulators prevent cultural capture when the few experts qualified to audit frontier AI already work for tech giants?