House Lawmakers Release Discussion Draft of Federal AI Governance Framework

Key Takeaways

On June 4, Reps. Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.) released a discussion draft of the Great American AI Act, a legislative proposal addressing various policy issues pertaining to artificial intelligence (AI) technology. The draft seeks public feedback before formal introduction and aims to balance AI innovation with safeguards for national security, workforce impacts, transparency and accountability. The legislation comes on the heels of a discussion draft released earlier in the year by Sen. Marsha Blackburn (R-Tenn.) in the U.S. Senate seeking to achieve similar outcomes. Both bills have not yet been formally introduced or advanced into markup.

What is in the discussion draft?

The proposal would create requirements for audits of advanced AI systems, establish workforce impact assessments, launch cybersecurity initiatives and support several research and development programs. Additionally, there are several provisions of importance for county review, including the below:

• Federal preemption of state laws on AI model development: The draft would temporarily preempt state and local laws specifically regulating the development of AI models for three years, while preserving the authority of governments to regulate AI use and deployment and enforce generally applicable laws. The bill would also call on the Government Accountability Office (GAO) to submit a report identifying federal statutes and regulations that affect AI innovation or unduly burden of AI infrastructure. 
  • Counties should monitor this provision closely as it could affect local policymaking efforts.
• New initiatives for workforce planning and skill building: The U.S. Department of Labor (DOL) would establish an AI Workforce Research Hub to analyze AI's impact on the workforce on a recurring basis, amend the  Worker Adjustment and Retraining Notification (WARN) Act (P.L. 100-379) obligations for employers to disclose when AI is a major factor in a qualifying mass worker layoff and study the creation of a Rapid AI Adjustment Assistance Program
  • Federal study of AI-driven workforce impact could inform county resource allocation to support workforce transition locally, fund training initiatives geared towards AI skill-building and support residents in affected sectors.
• Cybersecurity and open-source software: The bill would reauthorize the Cybersecurity Information Sharing Act of 2015 (CISA 2015) (P.L. 114-113), a landmark bill passed to increase the protected voluntary exchange of cybersecurity information between the private sector and federal government to improve the cybersecurity posture of the nation. The bill would also create grants through the Cybersecurity and Infrastructure Security Agency (CISA) to improve the security of open-source software. 
  • Counties support preventing the lapse of CISA 2015, and a study at the federal level on how to improve the security of open-source models as a means to utilize such low-cost tools for select, low-risk county operations.  

What are the next steps?

As the sponsors of the legislation solicit feedback and refine the legislation, counties should share feedback on the legislation with NACo for comment. NACo will continue to monitor and provide updates as needed.