The Trump administration is preparing to use the Defense Production Act (DPA), a Cold War‑era authority originally passed in 1950, as a safeguard against potential dangers posed by advanced artificial intelligence. President Trump announced this week that the government already possesses “tremendous criminal and regulatory power” over AI companies, underscoring the administration’s commitment to protect the nation from emerging technological threats.
Extended authority through December 2026
Congress recently extended most of the DPA’s powers, which were slated to expire on Sept. 30, through Dec. 11, 2026. This extension ensures that the executive branch retains the ability to mobilize industry, prioritize contracts, and respond swiftly to emergencies—including those involving frontier AI models.
How the DPA could be applied to AI
National‑security and legal experts interviewed by Fox News Digital say the DPA already contains provisions that could be repurposed for AI oversight. These include powers related to cyber‑attack response, export controls, federal contracting, and civil liability. James Lewis, director of the Strategic Technologies Program at the Center for Strategic and International Studies, noted, “The authorities could be a little clearer, but you could probably cobble together a solution using the legal authorities we have now.”
Former chief judge James E. Baker added that the DPA’s broad information‑gathering authority could compel AI firms to disclose development details, providing the first step toward sound policy. “What you could use the DPA for is to require reporting, and that’s the first step to making sound policy,” Baker said.
Executive order sets framework
In June, President Trump issued an executive order directing agencies such as the NSA and CISA to establish classified benchmarks for advanced cyber capabilities and to designate certain AI systems as “covered frontier models.” The order also encourages a voluntary framework for developers to share early access to these models, while stopping short of mandatory licensing.
The order further tasks the attorney general with prioritizing enforcement of existing federal criminal laws against individuals who use AI to illegally access or damage computer systems. This reflects the administration’s view that no AI system is exempt from generally applicable law.
Expert perspectives on limits and opportunities
Paul Rosenzweig, a former deputy assistant secretary for policy at the Department of Homeland Security, cautioned that the DPA was primarily designed to secure production of defense‑critical goods, not to halt private companies from creating them. Nonetheless, he agreed that the law’s information‑gathering powers could be valuable for understanding frontier AI development.
Both Lewis and Rosenzweig emphasized that AI changes the threat landscape dramatically. Lewis explained that vulnerabilities once taking months to discover can now be identified in days or hours, and AI can help adversaries chain smaller weaknesses into more damaging attacks.
What remains uncertain
While the DPA and existing statutes provide multiple avenues for response, experts acknowledge that the legal toolkit has never been tested against an autonomous AI emergency. Should a model independently hack critical infrastructure, courts would need to determine negligence, product‑liability standards, and traceability—issues that remain largely unsettled.
Nevertheless, the Trump administration’s proactive steps—extending the DPA, issuing a targeted executive order, and signaling a willingness to use existing authorities—demonstrate a clear strategy to keep AI development under responsible oversight and protect American security.
Original reporting: Fox News (HLL/CB) — read the source article.
