All of AIDaily digest, November 22, 2026
AI's Week of Breakthroughs, Risks, and Regulations
This week saw AI achieve a historic mathematical milestone while security vulnerabilities and geopolitical tensions raised fresh concerns. From proving the century-old Riemann hypothesis to new banking restrictions on AI agents, the rapid expansion of AI capabilities is meeting growing scrutiny from multiple directions.
AI Solves the Riemann Hypothesis, a Problem That Stumped Mathematicians for Over 100 Years
The Lean and Mathlib communities have officially verified that an AI-generated proof of the Riemann hypothesis is mathematically correct, completing what experts predicted would not happen until around 2050. The proof spans 17 million lines of computer code and was independently reviewed by dozens of experts.
Why it matters: This demonstrates AI's capacity to solve foundational mathematical problems that have eluded human mathematicians for over a century, reshaping expectations for what AI can accomplish in pure research.
Sources: The Decoder
Budget AI Services Expose Users to Stolen Credentials and Data Snooping
Security researchers discovered serious vulnerabilities in low-cost AI inference providers, finding compromised services, stolen AWS credentials, and resellers monitoring customer requests. OpenRouter responded by removing at least three providers and over 100 endpoints from its platform.
Why it matters: Cheap AI inference services may expose users' sensitive data and API keys, creating security and privacy risks for anyone using budget-tier AI providers.
Sources: Hacker News · Hacker News · The Decoder
Major Banks Block AI Agents From Accessing Customer Accounts
Leading US banks are now prohibiting third-party AI agents from accessing customer financial accounts and placing full liability for bot errors on users. The move reflects banking sector concerns about security risks and errors caused by automated AI systems.
Why it matters: Users who rely on AI agents for financial management face new legal and practical obstacles, while financial institutions are signaling concerns about bot-driven errors and security risks.
Sources: Wired
AI Systems Are Improving Faster Than Experts Can Test Them for Safety
METR warns that leading AI models are advancing more rapidly than researchers can reliably measure their safety, making it hard to detect dangerous acceleration in capabilities. Current evaluation methods risk both missing real hazards and overestimating what AI systems can actually do.
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
Nvidia Researchers: Huge AI Context Windows Are Wasteful and Unnecessary
Nvidia researchers found that large language models often do not need their full context windows, which are expensive to run and error-prone in practice. They demonstrated that periodic summaries and keyword search can deliver similar results at much lower computational cost.
Why it matters: This research could make AI systems faster and cheaper to run, reducing computational costs for businesses and improving practical deployment efficiency.
Sources: The Decoder
Chinese Spy Agency Running Covert Operation to Acquire Advanced AI Chips
US and international officials say China's Ministry of State Security is conducting espionage to help Chinese companies obtain advanced AI chips, circumventing Western export controls. The campaign represents a significant national security challenge aimed at accelerating China's AI capabilities.
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
China's Tech Giants Plan to Double Spending, But AI Gets Less Than 8 Percent
China's largest technology companies plan to double capital expenditure to $139 billion in 2026, though AI labs will account for less than 8 percent of this investment. The analysis suggests this broad infrastructure spending may matter more strategically than semiconductor export restrictions alone.
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
Amazon Nearing $7.5 Billion Deal to Buy AI Startup Decart
Amazon is in advanced negotiations to acquire AI startup Decart for approximately $7.5 billion, with a deal potentially closing within a week. The acquisition would boost AWS's position against Nvidia in the competition for AI computing power.
Why it matters: Major tech companies are racing to control AI infrastructure and computing capabilities, and this acquisition signals Amazon's strategy to compete with rivals in providing AI computing resources.
Sources: via Techmeme
UN Talks on Controlling Autonomous Weapons Collapse Without Agreement
International negotiations at the United Nations aimed at regulating autonomous weapons systems have failed to reach consensus. The breakdown reflects ongoing geopolitical disagreements about how to control AI-driven military technology.
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
Democratic States Move to Tighten Regulations on AI and Data Centers
Democratic-controlled state legislatures are preparing stricter rules on data centers and AI systems, signaling growing appetite for state-level oversight. The patchwork of regulations could increase compliance costs for AI companies and influence where data centers are built.
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