US Presses Meta to Agree to AI Reviews Amid Security Concerns
The Trump administration is urging Meta to submit its artificial intelligence models for voluntary government review, marking an escalation in Washington’s scrutiny of advanced AI systems amid growing national security concerns.
The request, conveyed via email to the social media giant, comes as the administration seeks to evaluate the capabilities and vulnerabilities of cutting-edge AI models. Meta, which launched its Meta AI model in April, remains the only major U.S. developer of artificial intelligence technology that has not agreed to voluntarily share its models with the federal government for assessment.
In a statement to Reuters, a Meta spokesperson said: “We share the administration’s goal of advancing U.S. leadership on robust and secure frontier AI. While we are working through the details, we hope to sign the agreement soon.” The company did not elaborate on the timeline or specific terms of the potential agreement.
The U.S. Commerce Department did not immediately respond to a request for comment on the matter.
This development follows earlier actions by the Biden administration, which in June ordered AI startup Anthropic to restrict access to its most advanced AI models for foreign nationals, citing national security risks. The move underscores a bipartisan concern in Washington about the potential misuse of powerful AI systems by adversaries.
Other leading AI developers have already engaged with the government on similar initiatives. OpenAI and Anthropic have been collaborating with U.S. officials to test unreleased AI models, while Google DeepMind, Microsoft, and xAI agreed in May to provide the government with early access to new models for national security evaluations.
According to sources familiar with the request, the Trump administration’s initiative is part of a broader effort to establish a voluntary framework for AI developers to submit their “frontier models” for government review before release to partners. This aligns with an executive order signed by President Donald Trump on June 2, which established a voluntary program for AI companies to offer their models for up to 30 days of evaluation.
National security officials argue that early access to advanced AI systems allows them to identify potential threats ranging from cybersecurity vulnerabilities to potential for AI-driven systems to enhance their own capabilities. The initiative reflects a growing recognition that artificial intelligence, while promising significant economic and scientific benefits, also poses novel risks that require proactive oversight.
Industry analysts note that the U.S. government’s push for greater transparency in AI development could set a precedent for how democratic societies balance innovation with security in the age of artificial intelligence, a topic of growing relevance as evidenced by recent market reactions to AI spending concerns (see related analysis).
As the race to develop ever more powerful AI systems accelerates, the interaction between Silicon Valley and Washington will likely play a crucial role in shaping the future of artificial intelligence governance. The current talks with Meta represent a significant test of whether the industry can embrace collaborative approaches to addressing the complex challenges posed by advanced AI.
Image Source: MYJOYONLINE
