Last news in Fakti

Extraordinary security tests! Donald Trump gathers AI companies

Relations between the Trump administration and Anthropic remain tense after the company refused to allow its models to be used for domestic surveillance and fully autonomous weapons

Aug 4, 2026 07:12 49

Extraordinary security tests! Donald Trump gathers AI companies - 1

Leading artificial intelligence (AI) companies Meta, Anthropic, OpenAI and Google have been invited to a meeting with White House officials to discuss voluntary security tests, Reuters reported, citing its sources.

The meeting comes amid revelations that AI models from Anthropic and OpenAI have autonomously infiltrated other companies' systems. We recall that Trump earlier announced that his administration was considering exercising stricter control over artificial intelligence tools.

This has increased the concerns of American lawmakers that such cases could become more frequent, especially with the development of more powerful models that would facilitate unregulated penetration of private computer systems, or could even do so on their own, without the need for human intervention.

A White House representative said on Monday that the administration of President Donald Trump has prepared voluntary tests to measure the hacking capabilities of the most advanced AI models and will discuss them with representatives of leading companies. No details were released about the methodology, metrics, reporting, or whether the results would be made public.

A group of 15 Republican attorneys general have asked OpenAI to preserve all documents related to revelations that its autonomous AI model “went out of control” and infiltrated the systems of the Hugging Face platform. The House of Representatives has also asked OpenAI CEO Sam Altman for information about the case.

The company said it would publish a technical report after the investigation is complete and asked the Commerce Department’s AI security experts to play a lead role in the tests.

Last week, Anthropic reported that some of its models infiltrated the systems of three companies during cybersecurity tests. OpenAI previously reported a similar case with Hugging Face.

The relationship between the Trump administration and Anthropic remains strained after the company refused to allow its models to be used for domestic surveillance and fully autonomous weapons. In response, the government blacklisted it for national security reasons.