Congress Urged to Establish Federal AI Framework Before Economic and Security Harms Escalate
A Massachusetts member of the House Energy and Commerce Committee is calling on Congress to pass a federal framework governing artificial intelligence, arguing that the window for action is narrowing and that inaction risks repeating the economic damage wrought on workers by decades of offshoring and automation.
Why It Matters
No federal statute currently governs how the most powerful AI systems are built, tested, or deployed. No independent body verifies the safety claims of leading AI developers, and no federal agency holds clear authority to intervene when something goes wrong. The absence of a legal structure, the lawmaker argues, leaves American workers, national security, and critical cyber infrastructure exposed to accelerating risks.
The concern is grounded in recent labor market signals: new and recent college graduates are experiencing higher unemployment than the broader workforce, and workers across the economy are questioning whether their current jobs will exist within five years.
What Happened
The Massachusetts congressman disclosed that he has been in bipartisan negotiations with Rep. Jay Obernolte (R-CA), his Republican counterpart on the House Energy and Commerce Committee, to draft a federal AI governance framework. The congressman described those talks as difficult and unresolved but said they have revealed broader bipartisan agreement on core principles than is commonly assumed.
The push gained urgency after AI developer Anthropic unveiled a system called Mythos, a model capable of identifying thousands of security vulnerabilities across major operating systems and web browsers. The company determined the system was too dangerous to release publicly, but the episode underscored how rapidly AI capabilities are advancing — and how far governance has fallen behind.
The congressman drew a direct comparison to the 2020 closure of a Brooks Brothers manufacturing facility in Haverhill, Massachusetts, which ended careers workers had built over decades. He argued that Congress failed those workers by not anticipating the disruption, and that a similar failure on AI policy is unfolding in real time.
By the Numbers
- 0 — Current number of federal laws governing the development or deployment of frontier AI systems
- 2020 — Year the Brooks Brothers factory in Haverhill closed, displacing long-tenured manufacturing workers
- Thousands — Number of software vulnerabilities the Anthropic Mythos model was reported capable of identifying across major platforms
- 2 — Number of House committee members — one Democrat, one Republican — currently engaged in bipartisan AI framework negotiations
The Proposed Framework
The congressman outlined four broad pillars he believes can attract bipartisan support. First, the largest AI developers — referred to in the industry as “frontier” labs — would be required to publish and adhere to safety and security frameworks, submit to third-party audits, and face penalties enforced by federal and state regulators. Whistleblower protections for employees at those companies would also be strengthened.
Second, the federal government would accredit private organizations to embed within frontier labs, assess safety practices, and trigger enforcement when companies fall short — rather than attempting to audit every AI model directly.
Third, employers would be required to disclose when AI is a direct cause of mass layoffs, updating existing Worker Adjustment and Retraining Notification Act requirements so that labor market disruptions driven by AI are tracked and reported in real time — giving policymakers earlier warning than they had during previous waves of automation and offshoring.
Fourth, the framework would allocate funding to developers maintaining open-source software that AI systems rely on, require frontier AI companies to provide those maintainers access to their models for security research, and reauthorize information-sharing agreements that allow companies to flag cybersecurity threats to one another without antitrust liability.
Zoom Out
Several states have moved to fill the regulatory vacuum, passing their own AI oversight measures. The congressman acknowledged that state-level work has been substantive but argued that a federal framework should incorporate the strongest state provisions while concentrating oversight at the frontier model development level — preserving states’ ability to address harms from AI systems already in use. Energy and technology affordability debates have increasingly intersected with AI infrastructure questions as data center demand surges nationally.
What’s Next
The bipartisan negotiations between the two committee members remain ongoing and unresolved. No legislation has been introduced. The congressman did not provide a timeline for when a formal bill might be ready for committee consideration, but characterized the current moment as a narrow window that is closing. Massachusetts Gov. Maura Healey, who faces her own political headwinds at home, has not been named as a participant in the federal discussions.