CAIP Commends the Release of the Bipartisan AI Task Force’s Landmark Report

The Center for AI Policy commends the release of the Bipartisan Artificial Intelligence Task Force's landmark report “Representatives Jay Obernolte (R-CA-23) and Ted Lieu (D-CA-36) deserve recognition for their leadership in bringing this report forward. This bipartisan effort acknowledges the need for a thoughtful, risk-based approach to AI governance. However, more work will be needed to develop appropriate laws.”

Washington, DC— Today, the US House of Representatives Bipartisan Artificial Intelligence Task Force released its long-anticipated AI policy report.

In response to the release of the report, Jason Green-Lowe, executive director of the Center for AI Policy (CAIP), a nonpartisan research organization dedicated to mitigating the catastrophic risks of AI through policy development and advocacy, released the following statement:

"The Center for AI Policy commends the release of the Bipartisan Artificial Intelligence Task Force's landmark report. This wide-ranging report marks a critical step forward in Congressional oversight of artificial intelligence (AI). It consistently points out the need to mitigate the risks posed by AI and suggests several helpful pathways for doing so, including further investment in AI standards, evaluations, and research."

CAIP agrees with the report’s authors that ‘a thoughtful, risk-based approach to AI governance can promote innovation rather than stifle it.’

CAIP maintains that robust safety measures are not obstacles to innovation and American technological leadership but essential catalysts for sustainable progress.“Likewise, we are pleased to see the report acknowledge that some risks are ‘truly new for AI due to capabilities that did not previously exist’ and that in some cases ‘when an AI issue has emerged recently…we need to more thoroughly consider how well existing regulatory regimes address that issue.’"Drawing from extensive stakeholder engagement and thorough analysis, the report represents a significant achievement in developing a consensus-driven approach to one of the world's most transformative technological developments.

CAIP supports the following Bipartisan House Task Force on Artificial Intelligence recommendations:

• Focus on responsible AI deployment
• Growing attention to transparency and explainability requirements
• Data quality and security are paramount for AI implementation
• Need for robust governance frameworks and human oversight
• Emphasis on bias detection and mitigation in AI systems
• Support the National Institute of Standards & Technology (NIST) in developing and operationalizing AI standards
• Appropriately counter the growing harm of AI-created deepfakes
• Meaningful human control over nuclear weapons

However, CAIP expresses serious reservations regarding the Task Force’s neglect of catastrophic risks. In its 273-page report, the Task Force used “catastrophic” only once in a footnote. The Task Force’s discussion of chemical, biological, radiological, and nuclear (CBRN) risks implicitly assumes that AI will never be able to effectively coach rogue states or terrorists on how to construct CBRN weapons. The report does not discuss the risk of advanced AI agents escaping human control. Because of risks like these, leading computer scientists, the leaders of major AI companies, and bipartisan majorities of the American public share a concern that without additional guardrails, increasingly advanced AI will be extremely dangerous to the future of humanity. The Task Force’s report did not effectively address this concern.

CAIP believes that to mitigate these catastrophic risks, the United States government needs to create a new federal agency called the Frontier Artificial Intelligence Administration.

Like other federal agencies, CAIP envisions that the Frontier AI Administration will be able to publish its regulations to ensure safe and responsible AI happens in America.

Although the Task Force’s report acknowledges in its introduction that some AI threats may require new regulatory frameworks, the body of the report does not contain enough detail about how or when these frameworks will be created. Congress has deliberated on AI for two years now, and it is time to start moving forward with decisive action.

Representatives Jay Obernolte (R-CA-23) and Ted Lieu (D-CA-36) deserve recognition for their leadership in bringing this report forward. This bipartisan effort acknowledges the need for a thoughtful, risk-based approach to AI governance. However, more work will be needed to develop appropriate laws. CAIP appreciates the Bipartisan Artificial Intelligence Task Force's thoughtful approach to developing these recommendations. We remain committed to collaborating with the new Congress to advance responsible AI development that serves all Americans.

The Center for AI Policy (CAIP) is a nonpartisan research organization dedicated to mitigating the catastrophic risks of AI through policy development and advocacy. Based in Washington, DC, CAIP works to ensure AI is developed and implemented with effective safety standards.

Learn more at centeraipolicy.org.