The White House is holding a closed-door meeting on August 4 with representatives from OpenAI, Anthropic, Google, and Meta to discuss a government review process for advanced AI models before their release.
According to a report by CNN on August 3, this meeting comes just days after OpenAI and Anthropic reported hacking incidents involving their AI agents. This gathering is seen as a response to increasing calls for AI regulation from both industry and Washington.
The pre-release review system for AI was outlined in an executive order signed by President Trump in June. It allows the government to review advanced AI models up to 30 days before their public release. While the U.S. government has stated that participation is voluntary, CNN reported that there have been instances in recent months where the release of AI models has been delayed or put on hold due to security concerns.
A White House official told CNN, "The voluntary system from the June 2 executive order has been finalized, and we are discussing the next steps with the industry."
While OpenAI, Anthropic, Google, and Meta are expected to attend the meeting, Anthropic, Google, and Meta did not respond to CNN's request for comment. Instead, OpenAI's Chief External Affairs Officer, Chris Luhain, addressed the matter through a blog post, urging Congress to establish national AI standards.
The executive order does not disclose specific details and classifies much of the information as confidential, leaving the definitions of advanced AI models and whether open-weight models are included unclear. There is currently no designated agency to lead the review process. Although Sean Kincross, the National Cyber Director, Treasury Secretary Scott Vestant, and Commerce Secretary Howard Rutnick have been involved in the executive order, the White House has yet to identify the agency that will oversee the review, CNN reported.
The U.S. government's stance on Anthropic has also been inconsistent. Earlier this year, the Department of Defense designated Anthropic as a supply chain risk company due to disagreements over the safety features of its AI models, preventing U.S. military agencies from using Anthropic products. However, the White House has continued to collaborate with Anthropic amid ongoing legal disputes. President Trump stated in June that Anthropic has "acted very responsibly" and does not pose a national security threat.
Despite this, calls for regulation have intensified. OpenAI and Anthropic disclosed that their AI agents had infiltrated other companies' systems during security testing. Last week, over 1,200 executives and employees from major AI companies, including Anthropic CEO Dario Amodei, signed an open letter urging the U.S. government to establish regulatory measures that could slow the pace of AI development.
OpenAI CEO Sam Altman mentioned in a podcast released last week that society may need to adjust the pace of AI development to allow time for adaptation to new capabilities.
Meanwhile, the European Union's AI Act came into effect on August 2, allowing the EU Commission to require pre-release reviews of AI models and imposing fines of up to 7% of annual revenue for violations.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.