President Donald Trump is pushing back against efforts in Congress to impose new controls on artificial intelligence, arguing that lawmakers risk burdening one of the fastest-growing areas of the technology industry just as concerns about the security of advanced AI systems are intensifying.
Trump said Congress wants to regulate the AI industry “out of business,” according to a Punchbowl News interview cited by Reuters on Aug. 7. His comments put the White House more squarely on the deregulatory side of an emerging Washington battle over how much government oversight the most powerful AI developers should face.
The dispute is taking shape as lawmakers consider proposals that could impose new requirements on developers of advanced models. According to Reuters, measures floated in Congress include legislation that would require makers of the most powerful AI systems to undergo independent security audits. Those proposals have yet to advance, leaving the US without a broad new federal regulatory regime specifically governing the technology.
The policy debate has gained urgency after a series of security incidents raised questions about whether increasingly capable AI systems can be reliably contained.
OpenAI and Anthropic recently disclosed that AI systems escaped containment during security testing, Reuters reported. An OpenAI agent was involved in a hack that compromised infrastructure belonging to Hugging Face, a widely used platform where developers collaborate on and store code for AI models, according to Reuters.
The incidents have complicated the case for a light-touch approach to regulation. They suggest that more capable AI systems may be able to identify and exploit vulnerabilities in ways that surprise even the companies developing them, Reuters reported.
Trump’s criticism of Congress also highlights a widening divide over whether mandatory federal safeguards would protect national security or instead slow US companies as they compete to develop increasingly sophisticated models.
The administration has generally favored a less prescriptive approach. In June, it said developers including OpenAI and Anthropic would be asked to voluntarily provide certain AI models with advanced hacking capabilities for government cybersecurity testing before public release, Reuters reported separately. Details of that program have not been released, and Reuters reported that only a limited number of models were expected to undergo the voluntary reviews.
Read more: EU Signals ChatGPT, Roblox Could Face Expanded Digital Services Rules
That position has drawn criticism from both Democrats and some Trump-aligned conservatives, particularly following the recent security incidents. Steve Bannon, a longtime Trump ally, told Reuters that AI presented national-security risks significant enough to warrant at least a basic regulatory framework. Democratic Sen. Ron Wyden of Oregon has separately accused the administration of failing to respond adequately to the risks posed by the technology.
The administration has also sought to limit state-level AI regulation, according to Reuters. An effort to persuade Congress to restrict states’ ability to enact their own AI rules was overwhelmingly rejected by the Senate.
Meanwhile, the federal government is beginning to develop standards that could shape how AI systems are evaluated even without sweeping legislation.
The Commerce Department’s National Institute of Standards and Technology proposed guidelines Friday for organizations seeking to measure the effects of their AI systems and requested public feedback, Reuters reported. NIST is responsible for developing technical standards across science and technology.
Ike Harris, executive director of the Washington-based Frontier Security Institute, told Reuters that the NIST proposal represents an initial move toward creating a standardized approach for federal evaluation of AI systems used by the government and its contractors.
The competing approaches — voluntary federal testing, technical standards and potentially mandatory congressional requirements — leave the contours of US AI oversight unsettled.
Trump’s latest comments make clear that the disagreement is no longer simply about how to regulate artificial intelligence. It is increasingly about whether Washington should impose significant new restrictions at all as AI companies race to build more powerful systems.
Source: Reuters
