Reuters: Trump administration urged Meta to submit AI models for voluntary review

Meta Platforms was reportedly asked by the Trump administration to submit AI models for a “voluntary review,” a step that underscores how US policymakers are trying to increase oversight of advanced artificial intelligence systems. Reuters reported on June 23 that, according to a report from The New York Times published Tuesday, the administration pushed Meta to take part in the process.

The reporting described the request as part of a broader effort to encourage companies to share information about how their AI models behave, rather than requiring companies to do so through a formal rule. Reuters did not say in the Yahoo Finance recap what specific models were at issue or what would be measured in any review, and Meta was not described as making a detailed public response in the account summarized by Reuters.

In practice, voluntary reviews are often designed to reduce uncertainty for regulators and the public by bringing external scrutiny earlier in a model’s lifecycle. For technology companies, participation can also become a reputational announcement, but the tradeoff is time, cost, and the potential exposure of information that a company might otherwise treat as proprietary.

The episode matters for Meta because the company’s AI systems are central to its products and the way it targets content and ads. Oversight efforts focused on model behavior and risk controls therefore land directly on the infrastructure that supports those capabilities. Even when the process is voluntary, a high-profile company like Meta can end up setting expectations for how other firms interpret what regulators want to see.

For markets, the immediate question is not only whether Meta complied, but what the company disclosed, if anything, about its approach to model safety and evaluation. Reuters, as reflected in the Yahoo Finance item, did not include details on the scope, timing, or outcomes of the review request, limiting what investors can infer about operational changes inside Meta.

Still, the request highlights a policy environment in which AI oversight is moving toward concrete, model-level scrutiny. With different administrations and agencies potentially favoring different mechanisms, voluntary processes may become a recurring theme, especially if policymakers see them as faster to implement than legislation or rulemaking.

Looking ahead, what to watch is whether Meta publicly addresses any aspect of the request, including whether it agreed to participate and whether the company will describe any new evaluation practices. It will also be important to see if more companies announce similar participation, or if regulators publish guidance that makes the criteria for voluntary review clearer.