
Nov 18, 2026 · 17 min
Quiet AI safety ties bridge Washington and Beijing
Inside the AI labs’ secretive U.S.-China safety meetings
As official U.S.–China dialogue stalls, informal contact among AI labs may offer rare visibility into how each side manages frontier-system risks.
- 1Concordia AI quietly connects China’s safety community with OpenAI, Google DeepMind, and other U.S. frontier labs.
- 2Technical cooperation persists despite strategic rivalry, research-security concerns, and limited progress through official government channels.
- 3Anthropic’s expected IPO adds a market subplot, with research warning that large offerings often lag broader markets after debut.
Don't miss
The episode’s clearest tension emerges as informal lab-to-lab contact appears more productive than official U.S.–China safety diplomacy.
The brief
Concordia AI operates at the intersection of consulting, research, evaluation, and convening, linking China’s AI safety ecosystem with U.S. labs and policymakers.
The organization recently visited OpenAI and has engaged with Google DeepMind, illustrating a quiet channel that remains open despite intensifying U.S.–China rivalry.
Government efforts, including a proposed AI safety hotline and discussions around Trump and Xi, have produced limited results; technical communities still value direct visibility.
American labs and safety advocates face a difficult tradeoff: preserve communication and pursue global pacing while managing research-security risks and conflicting U.S. policy signals.
The episode closes with a preview of Anthropic’s expected IPO, shifting from geopolitical risk to investor demand and the weak post-debut record of large offerings.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.