Trump: AI Companies Pledge Stricter Self-Oversight, Legislation Appears Off the Table
President Trump reports that American AI companies have committed to better oversight of their models through internal audits and external assessments. The voluntary approach appears to further reduce the likelihood of binding federal AI legislation in the US.
Voluntary agreements instead of legislation
US President Donald Trump has announced that AI companies have made commitments to better oversee their models. According to reporting by NU.nl Tech, this involves agreements intended to create more room for internal checks, also known as audits, and for external assessments of AI systems. Which companies exactly are involved has not been confirmed based on the available reporting.
The gist of the report is that these voluntary commitments appear to further reduce the need for binding, national AI legislation in the United States. This continues a trend that has been visible for some time: self-regulation by the industry rather than legal obligations imposed by the government.
No federal AI law in the US
The United States currently has no comprehensive federal law regulating AI models. Regulation so far has proceeded through sector-specific rules, presidential executive orders, and agreements the government makes with individual companies. This fragmented approach stands in stark contrast to, for example, the European Union, which has established an extensive legal framework with the AI Act.
In January 2025, Trump revoked an executive order from his predecessor Joe Biden that focused on AI safety. At the time, he advocated for a less restrictive, innovation-friendly approach to the sector. The commitments now reported fit this pattern: preferring voluntary agreements with companies over legal obligations that could slow down innovation.
Voluntary versus enforceable
An important distinction is that the reported commitments are voluntary and not legally enforceable, unlike legislation. Companies can, in theory, withdraw from such agreements without facing direct legal consequences, whereas violating a law can lead to sanctions.
This type of arrangement is reminiscent of an earlier episode in 2023, when, under the Biden administration, a series of major AI companies made voluntary safety commitments at the White House. Whether the commitments now reported are comparable in content and scope cannot be determined based on the available reporting.
What remains unclear
There is no clarity, based on current reporting, about the exact date and location of Trump's statements, any quotes, or the precise details of the announced internal checks and external assessments. It is also unknown who the external assessors would be or what standards would be applied.
Outlook
The development underscores that the US approach to AI regulation currently relies on voluntary cooperation from the industry itself, rather than on binding legislation. For companies and regulators worldwide, the question remains how effective such non-enforceable commitments are in practice, especially now that other regions, such as the EU, apply a stricter legal framework.