The White House says Adobe, Cohere, Microsoft, Anthropic, OpenAI, and Common Crawl made voluntary commitments to combat nonconsensual image deepfakes and CSAM
The White House has announced that several major AI vendors, including OpenAI and Microsoft, have committed to taking steps …
Context & Ripple Effects
This is a more targeted extension of the White House’s earlier voluntary AI-governance drive, which began with commitments on cybersecurity and watermarking from leading model providers in 2023 the initial White House AI governance pledges.
It follows the administration’s May call for companies to address the creation, distribution, and monetization of nonconsensual sexual AI imagery a call focused on sexual-image abuse, while broadening the signatory set to include Common Crawl alongside model and software vendors.
First-order effects
- Adobe, Cohere, Microsoft, Anthropic, OpenAI, and Common Crawl are publicly accountable for voluntary measures aimed at nonconsensual image deepfakes and CSAM.
- The commitments give the White House a shared industry response on two concrete abuse categories without establishing a mandatory rule.
Second-order effects
- Other AI providers and data intermediaries face greater pressure to demonstrate comparable safeguards, particularly after a separate industry accord addressed election-related deceptive deepfakes the earlier election-deepfake accord.
- Trust-and-safety, product, and data-governance teams become more central to how participating companies document and operationalize their AI safety posture.
Third-order effects
- If repeated across abuse categories, voluntary pledges could become a de facto baseline for state-facing AI governance, even though their effectiveness depends on the specific measures and follow-through companies disclose.
- The inclusion of a web-data organization points toward a wider enforcement surface spanning both AI model deployment and the information inputs that support the ecosystem.
The trend: The White House is using issue-specific voluntary commitments to extend AI governance from broad model-safety principles into distinct harmful-use and ecosystem risks.