Trump Signs Executive Order for Voluntary AI Model Review
Donald Trump signed an executive order to establish a voluntary framework allowing the federal government to review powerful new AI models prior to their public release. This order, announced Tuesday, marks an effort by the president to enhance oversight on cybersecurity and national security risks associated with artificial intelligence, signaling a shift from his previous deregulatory policies.
Voluntary Sharing of AI Models with Government
Under the new guidelines, technology companies will be requested to submit their AI models to the government for voluntary review up to 30 days before making them publicly available. The executive order states that this process is intended to bolster national security, especially in the realm of cybersecurity.
The order does not mandate compulsory review of AI models by tech firms, a feature that was reportedly considered in earlier drafts. Some of Trump's more stringent MAGA supporters had called for a more rigorous review process, while advocates within the tech industry favored a less restrictive approach.
Shift from Previous Low-Regulation AI Policies
This executive order further illustrates Trump's departure from his initial low-regulation stance on AI. Early in his presidency, he rescinded a Biden-era executive order that had set standards for the safe development of AI technologies.
Concerns Over AI Safety and Cybersecurity
The new regulatory measures come amid growing apprehension about the potential dangers posed by advanced AI models, particularly if misused. For instance, a model with sophisticated cybersecurity capabilities has alarmed AI safety experts, government officials, and technology companies due to its unprecedented ability to exploit vulnerabilities in widely used software.
Government Collaboration with Tech Companies
Last month, the Trump administration entered agreements with Microsoft, Google DeepMind, and xAI to review early versions of their AI models before public release. Although the federal government has since removed details of this agreement from its website for unclear reasons, the Center for AI Standards and Innovation (CAISI), part of the U.S. Department of Commerce, maintains similar arrangements with OpenAI and Anthropic. The government emphasizes that such information sharing is standard practice and vital for national security, though some free speech advocates caution that excessive government control might lead to censorship.
Roles of Federal Agencies and Workforce Expansion
The National Security Agency and the Department of Defense will assist in identifying which AI models require government evaluation, while the Treasury Department will focus on detecting vulnerabilities within AI systems. Additionally, the Trump administration has directed efforts to recruit more cybersecurity and AI professionals and to strengthen cybersecurity infrastructure at critical facilities such as hospitals and banks.
Additional AI-Related Executive Order
Trump also announced another AI-focused executive order aimed at preventing states from independently regulating AI technologies. This order established a federal task force to challenge state-level AI regulations.
