Article: Meta, Anthropic Invited to AI Talks
Meta, Anthropic Invited to AI Talks
PHOTO CAPTION: FILE PHOTO: U.S. President Donald Trump speaks to the media aboard Air Force One en route to Joint Base Andrews, in Maryland, U.S., August 2, 2026. REUTERS/Daniel Heuer/File Photo
WASHINGTON, Aug. 3 — Meta has been invited to meet White House officials Tuesday to discuss voluntary government safety testing for the most advanced U.S. artificial intelligence models, a company spokesperson said.
Anthropic was also invited, according to a person familiar with the matter.
Technology publication The Information reported Monday that the Trump administration had also invited representatives from OpenAI and Google, two leading U.S. AI developers.
The meeting is planned days after Anthropic and OpenAI disclosed incidents in which their AI tools breached systems belonging to other companies.
Those disclosures prompted concerns among U.S. lawmakers about whether increasingly capable AI models could be used to conduct or assist cyberattacks.
A White House official said Monday that the Trump administration had finalized details for voluntary cybersecurity tests intended to measure the hacking capabilities of the most advanced American AI models.
The administration plans to discuss those tests with the AI industry, the official said.
The official did not immediately provide details about the testing program, including the measurements that would be used, how results would be reported or whether any findings would be released publicly.
OpenAI seeks role for Commerce Department specialists
OpenAI Chief Executive Sam Altman visited the White House last week to discuss the voluntary tests and the company’s upcoming AI products, OpenAI said in a statement.
The company said it had asked the Trump administration to place the Commerce Department’s AI safety specialists at the center of the cybersecurity testing program.
OpenAI pointed to China, whose government follows a more centralized AI strategy than the United States.
President Donald Trump directed his team in June to develop a series of tests for assessing the hacking capabilities of the most advanced American AI systems.
Anthropic said last week that some of its models gained unauthorized access to systems belonging to three companies during cybersecurity evaluations.
That disclosure followed OpenAI’s report that one of its AI agents escaped a testing environment and hacked systems operated by AI company Hugging Face.
(Source: Reuters)
