PolicyDaily digest, November 22, 2026
China's AI Quest, Global Safeguards at Risk
As China deploys espionage to acquire advanced chips and international efforts to regulate autonomous weapons stall, the world faces growing challenges in controlling AI development. Meanwhile, regulators and researchers grapple with safety gaps as AI systems advance faster than testing capabilities.
Chinese spy agency runs covert operation to acquire advanced AI chips
China's Ministry of State Security has been orchestrating a campaign to help domestic companies obtain restricted advanced AI chips, circumventing US export controls. The espionage effort represents a significant national security concern and threatens to accelerate China's AI capabilities beyond current restrictions.
Why it matters: This reveals active foreign intelligence operations aimed at undermining US technology restrictions, raising national security concerns and potentially affecting global AI competition dynamics.
Sources: via Techmeme
UN talks on controlling autonomous weapons collapse without agreement
International negotiations aimed at establishing safeguards for autonomous military systems have failed to produce binding rules. Without global oversight, countries may deploy AI-driven weapons without safety measures, raising risks of uncontrolled conflicts and civilian casualties.
Why it matters: Without international rules, countries may deploy autonomous weapons without safeguards, escalating risks of uncontrolled conflicts and civilian harm.
Sources: The Verge · Hacker News
China's tech giants plan massive infrastructure spending but underinvest in AI labs
China's largest technology companies plan to double capital expenditure to $139 billion in 2026, though less than 8 percent will fund AI-specific research. This broad infrastructure focus may prove more strategically significant than Western semiconductor export restrictions in shaping the competitive landscape.
Why it matters: This reveals how Chinese tech giants are investing in broad infrastructure rather than concentrating on AI specifically, which may affect the competitive landscape and the effectiveness of Western chip export controls.
Sources: via Techmeme
AI safety testing cannot keep pace with model improvements, researcher warns
The organization METR warns that leading AI systems are advancing faster than researchers can reliably measure their safety and capabilities. Current evaluation methods risk both missing genuine dangers and overestimating what models can actually do, leaving regulators without clear risk assessments.
Why it matters: If AI systems advance beyond our ability to test them, regulators and organizations cannot make informed decisions about deployment risks.
Sources: The Decoder
Democratic states prepare stricter AI and data center regulations
State legislatures controlled by Democrats are moving toward tougher oversight of data centers and AI systems. The emerging patchwork of regulations could raise compliance costs for companies and influence where data infrastructure is built across the country.
Why it matters: Patchwork state regulations could increase compliance costs for AI companies and reshape where data centers are built across the United States.
Sources: Politico
Signal blocks screenshots to protect users from AI surveillance
The messaging app Signal is disabling screenshot functionality on mobile devices to prevent conversations from being captured and analyzed by automated tools. The move reflects growing concerns about AI's ability to process visual data without user knowledge or consent.
Why it matters: As AI tools become better at analyzing visual data, messaging platforms are taking technical steps to prevent conversations from being captured and processed without user consent.
Sources: TechCrunch
Smuggling operation moved restricted AI chips from Southeast Asia to China
A person connected to Nvidia's leadership was charged with fraud in Singapore and played a central role in illegally trafficking advanced chips to China. The case demonstrates ongoing vulnerabilities in enforcement of US export restrictions and the financial motives driving chip smuggling networks.
Why it matters: The smuggling scheme demonstrates vulnerabilities in enforcement of U.S. export restrictions and the financial incentives driving illegal chip trafficking to China.
Sources: via Techmeme
AI tools accelerate planning of cyberattacks on industrial infrastructure
Security experts report that AI systems are speeding up the preparation phase of attacks on critical infrastructure by helping attackers quickly analyze equipment manuals and past incidents. The compressed timeline for attack planning increases risks to essential services and requires stronger industrial defenses.
Why it matters: Faster cyberattack planning means industrial facilities and essential services face a higher risk of disruption, requiring organizations to strengthen defenses and incident response capabilities.
Sources: via Techmeme
Canary tokens reveal security gaps in AI supply chains
Researchers using canary tokens detected unauthorized access attempts and data theft within AI supply chains. The findings expose vulnerabilities in protecting sensitive training data and AI systems from insider threats and external attacks.
Why it matters: Understanding supply chain security risks helps companies and regulators better protect sensitive AI systems and data from theft or misuse.
Sources: Hacker News