New Bipartisan AI Safety Bill Will Move U.S. and China Beyond the “Red Phone”
WASHINGTON, D.C. – Today, Congressmen Sam Liccardo (D-CA) and Kevin Kiley (I-CA) will introduce legislation to establish U.S.-China coordination on frontier artificial intelligence safety.
The proposal advances that work on two tracks: clearing legal barriers so that top American experts in AI labs, companies, universities, think tanks, and other technical experts can engage directly with Chinese counterparts, while directing the State Department and the rest of the Administration to pursue negotiations with China on binding, verifiable safeguards. As Congressman Liccardo put it, “A lot of technical work must begin now, before political leadership can ever fruitfully meet at the negotiating table. Both nations will need to forge shared safety metrics, rigorous evaluation protocols, standardized testing, and critical verification mechanisms.” This proposal focuses on developing routine U.S.-Chinese communication at the technical level to prevent critical incidents from developing. “By establishing red lines now,” Liccardo added, ”we can avoid the need for red phones later.”
“The AI industry has focused on running faster, but no human wins a race to superintelligence that exceeds human control,” said Liccardo. “We don’t talk to the Chinese because we trust them or agree with them; we shouldn’t and don’t. We talk to the Chinese because—as with the USSR during the Cold War—all of our lives depend on it.”
The framework would:
- Pursue binding U.S.-China AI safety standards
- Direct the State Department to seek negotiations with China covering model testing and evaluation, independent auditing, transparency, incident reporting, capability pacing, and mechanisms to verify compliance.
- Protect American intellectual property.
- Direct the Commerce Department to pursue negotiations with China on protections against the compromise of U.S. intellectual property through model distillation.
- Target catastrophic AI risks.
- Focus negotiations on loss of control and alignment failures, AI-enabled chemical, biological, radiological, and nuclear (CBRN) threats, and major cybersecurity risks.
- Create an AI “red phone.”
- Establish a direct hotline between the U.S. and Chinese governments for urgent communication involving significant AI risks or incidents.
- Build common technical standards.
- Authorize $100 million for the Center for AI Standards and Innovation (CAISI) to bring together experts from industry, academia, and civil society and work with international counterparts on common testing, evaluation, and verification standards.
- Enable tightly controlled expert engagement.
- Allow U.S. AI safety experts to work with Chinese counterparts on narrowly defined safety standards while requiring advance registration with the State Department and permitting U.S. government observation. The framework prohibits participants from sharing proprietary intellectual property, information that could assist model distillation, or sensitive cybersecurity and national security information.
- Require congressional oversight.
- Require the State and Commerce Departments to provide Congress with monthly updates on progress.
Read the full framework here.