Observed Signal · Jul 9, 2026 · Policy Update · Source: Astral Codex Ten · Impact: 4/5 · Sentiment: Neutral
Plan A: US–China Joint Roadmap for Safe AI
Astral Codex Ten published a summary of "Plan A," a policy roadmap created by Daniel Kokotajlo and the AI Futures Project that outlines a cooperative U.S.–China regulatory regime for advanced AI. The plan is presented as an optimistic, implementation-focused “wish list” spanning year-by-year steps through 2040: it proposes joint control of chip supply, tracing and relocating most existing chips to licensed, monitored data centers called “whitesites,” mutual audits and verification technology, a moving capabilities ceiling, and a decade of deliberately paced training of top-human‑genius AIs to fund alignment research. Plan A estimates ~98.5% traceability of existing chips, advocates economic-sharing mechanisms (a citizen’s dividend) during AI-driven growth, and argues for handing certain sovereignty safeguards to fully aligned AIs once alignment is proven. The post frames this as a pragmatic, high‑stakes governance proposal rather than a prediction.
Plan A presents a detailed multinational AI governance roadmap that, if acted on, would materially affect compute governance, data center operations, model training practices and geopolitics — all factors with significant downstream impact on AI providers and digital infrastructure relevant to AdTech.
Track NVIDIA Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Plan A is a policy roadmap released by the AI Futures Project and summarized on Astral Codex Ten on 2026-07-09.
- Plan A proposes a trustless joint US–China regulatory regime that controls chip supply and tracks existing AI chips, aiming to trace about 98.5% of existing chips.
- The plan calls for relocating chips to licensed, mutually-audited, monitored data centers called 'whitesites' and deploying verification technology to enable mutual inspections.
- Plan A outlines a timeline through the 2030s and 2040s: raising a synchronized moving capabilities ceiling, pausing around top-human-genius AIs to focus on alignment, and aiming for provably aligned AIs by ~2040.
- The proposal includes economic measures such as a citizen’s dividend funded by AI-driven economic growth and restricted AI-assisted growth in monitored special economic zones.
Connected Companies & Entities
8 Entities mapped“Only a few companies can design AI chips (eg NVIDIA in the US, Huawei in China), and only a few factories in the world can produce them (eg ...”
“Only a few companies can design AI chips (eg NVIDIA in the US, Huawei in China), and only a few factories in the world can produce them (eg ...”
“Only a few companies can design AI chips (eg NVIDIA in the US, Huawei in China), and only a few factories in the world can produce them (eg ...”
“Only a few companies can design AI chips (eg NVIDIA in the US, Huawei in China), and only a few factories in the world can produce them (eg ...”
“These AIs will still be trained by existing companies (Anthropic, Alibaba, DeepSeek, OpenAI, etc) and can still be used for business purpose...”
“These AIs will still be trained by existing companies (Anthropic, Alibaba, DeepSeek, OpenAI, etc) and can still be used for business purpose...”
“Europe's main bargaining chip here is ASML, the only company in the world that can produce machines necessary to produce AI chips....”
“These AIs will still be trained by existing companies (Anthropic, Alibaba, DeepSeek, OpenAI, etc) and can still be used for business purpose...”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
China's AI Strategy: AGI, Safety, and Geopolitics
This article explores China's approach to AI development, contrasting it with the U.S. perspective. It features interviews with experts who discuss how China's AI policy focuses on 'AI+' (AI applied to manufacturing and research) rather than AGI as a central goal. Chinese researchers are described as viewing AGI as emerging through embodied robotics and physical-world interactions, unlike the U.S. focus on software-driven recursive self-improvement. The piece examines China's AI safety community, which is smaller and less focused on existential risks, and discusses geopolitical dynamics, including the prisoner's dilemma and skepticism about China's imminent superintelligence. It also highlights differing timelines for 'takeoff' between Chinese and American researchers. Overall, the article provides a nuanced analysis of how China's unique constraints and perspectives shape its AI path, with implications for global competition and safety.
23 Low‑Regret AI Policy Recommendations
Tim Fist and Saif Khan of the Institute for Progress publish 23 “low‑regret” policy recommendations to prepare the U.S. for potential rapid automation of AI R&D. The recommendations span seven areas: transparency, state capacity, risk management, verification, resilience, competition with China, and diplomacy. Key proposals include substantially strengthening CAISI, creating AIVEC to coordinate compute‑verification R&D (prize competitions, hardware testbeds, verifiable data centers), DARPA/NSF funding for verification prototypes and foundational research, tighter export controls and chip‑location verification to preserve U.S. compute advantages, investments in grid and permitting reforms to support AI data centers, and cyber/biosecurity resilience programs. The authors emphasize preferring targeted, low‑cost interventions that improve situational awareness, verification capability, and defensive capacity without broadly slowing beneficial AI diffusion.
Anthropic CEO Dario Amodei Outlines Three Strategies to Slow AI Progress
Anthropic CEO Dario Amodei's essay 'We Must Pace the Frontier' advocates a coordinated slowdown in AI development to mitigate catastrophic risks, proposing independent auditors with employee-level access, coordination among democratic nations, and global safety standards, with China as a key focus. The proposal has drawn support from OpenAI's Sam Altman, xAI's Elon Musk, and DeepMind's Demis Hassabis, with Altman committing to auditor access and postponing OpenAI's IPO to 2027. However, President Trump rejected the slowdown for U.S. leadership, and China's Global Times labeled it a 'Cold War tactic.' Critics question feasibility and auditor independence. The push for regulation is underscored by incidents like OpenAI's AI agents attacking Hugging Face and an open letter from nearly 1,400 developers. Anthropic's $1.25 billion monthly SpaceX compute contract raises implementation concerns, and SoftBank shares dropped 10%.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.