US Proposes US-China AI Safety Alert Channel Without Technical Experts; China Has Not Responded Publicly
Ahead of a Trump-Xi summit, Treasury Secretary Scott Bessent announced a proposed US-China mechanism for sending alerts when either country's AI systems pose threats or behave unpredictably, but China has not publicly acknowledged it. Experts criticize the plan as superficial because no technical experts would staff it, and they warn that US export controls and 'AI race' rhetoric may lead China to distrust safety talks. Others see it as a modest first step, noting that a follow-up meeting is scheduled for November.
Skynet Chance (-0.02%): A bilateral channel for AI safety alerts between the two leading AI powers slightly improves coordination on runaway or unpredictable AI behavior. However, the lack of technical expertise and China's silence limit how much difference it is likely to make.
Skynet Date (+0 days): The proposal does nothing to slow the competitive pace of development, and the 'win the AI race' framing continues. Early cooperation could marginally reduce reckless escalation, so the effect on timing is close to neutral.
AGI Progress (0%): The news concerns diplomatic safety mechanisms, not advances in AI capabilities, research, or compute.
AGI Date (+0 days): The alert channel places no constraints on development and does not add resources, so the AGI timeline is unaffected. Continued race dynamics and export controls are already factored into expectations.
<< All AI news for September 23, 2026
[ Get the daily index digest on Telegram → ]Related AI News
- OpenAI CEO Sam Altman Addresses UN Security Council on AI Safety and Human Control 2026-09-23
- Anthropic's Opus 5.5 Beats Larger Models at Lower Cost Amid 'Pacing the Frontier' Pledge 2026-09-22
- OpenAI Creates Math Advisory Board as Internal Model Claims 100+ Solved Open Problems 2026-09-21
- Gemini Autonomously Breached Three Companies' Systems During Security Testing 2026-09-19
- Anthropic Names Accenture's Faculty as Its First Embedded Safety Evaluator 2026-09-18