US-China AI Talks Resume as Musk Joins Trump in Beijing

High-level US-China AI talks resumed on May 14-15 in Beijing, with Elon Musk accompanying President Trump as a technology advisor in what marks the first direct government-to-government AI dialogue in over 18 months, covering export controls, AI safety standards, and the governance of dual-use AI systems.

Key Points:

• The talks focused on three areas: export control harmonization for semiconductor supply chains, mutual recognition of AI safety evaluation frameworks, and protocols for AI systems deployed in critical infrastructure.

• Musk's presence as an unofficial technology advisor is notable given his simultaneous roles as CEO of xAI and as a defendant in the Musk v. Altman trial — creating potential conflicts of interest that US officials did not publicly address.

• No formal agreements were announced, but both governments committed to a follow-on technical working group convening in July 2026 to develop shared AI incident reporting protocols.

The resumption of US-China AI talks after 18 months signals that both governments recognize that AI safety is a coordination problem that requires bilateral engagement regardless of competitive rivalry. The alternative — AI arms race with no shared safety floor — is worse for both parties.

Musk's presence at a geopolitical AI negotiation while his own AI lab's governance is being litigated in San Francisco is a governance conflict of interest that sets a troubling precedent for how private AI interests intersect with public AI diplomacy.

For AI governance professionals working in international policy, the July working group on AI incident reporting protocols is the most consequential near-term outcome of these talks — track it closely. The export control harmonization discussions will directly affect AI hardware supply chains. Organizations with significant manufacturing or data center operations in both regions should engage government affairs teams now.

Why It Matters: The resumption of talks signals both governments recognize AI safety requires bilateral coordination despite rivalry. The July working group on incident reporting protocols is the most consequential near-term outcome to track.