U.S. President Donald Trump's administration may review artificial intelligence (AI) models before they make it to broader users. The news comes amid growing cybersecurity concerns surrounding Anthropic’s new Claude Mythos model. This would be a major shift from the earlier approach, where the government largely allowed AI companies to develop and release models with minimal interference. 

Also read: Anthropic Mythos risks force public banks in India to revamp IT investments: Report

Government oversight of AI models

According to a New York Times report, the White House is considering creating an executive order to form a special group made up of technology leaders and government officials. The group will reportedly study and design possible rules or review systems for new AI models before they are released to the public. However, the discussions are underway, and no final decision has been announced yet.

As of now, the Trump administration is discussing the AI oversight ideas with executives from big tech companies like Google, Anthropic, and OpenAI. In addition, it is also planning to create a system that gets early or priority access to advanced AI models before the public can use them. The NYT reported that several White House officials are supporting the approach as a way to monitor potential risks linked to powerful AI systems like the Anthropic Claude Mythos.

Must read: BT explainer: Inside Claude Mythos, the AI is forcing a rethink of global cybersecurity

Well, this is not the first time the US government has considered oversight on AI models, as the Biden administration in 2023 also issued an executive order to provide safety test results. The motive for such rules will ensure that the AI tools people use daily are unbiased and secure to use.

However, Donald Trump has openly supported AI innovation, saying that “We can’t stop it with foolish rules and even stupid rules.” Whereas US Vice President JD Vance also highlighted that the “excessive regulation of the AI sector could kill a transformative industry just as it’s taking off.” But, now the growing concerns over powerful AI models, cybersecurity risks and national security implications appear to be pushing the government towards a more cautious approach.