AI Daily · Nov 22, 2026 · SunAI Daily
ICREAT AI NEWS
Anthropic mathematician claims two independent proofs of the Riemann Hypothesis with Claude's help
Anthropic mathematician Levent Alpöge claims to have proven the Riemann Hypothesis and its generalized form with Claude's assistance, releasing only about 1.3GB of Lean code rather than a full paper, corresponding to roughly 250,000 pages.
Product launches/updates
Qwen4 Max launches on Together AI, fixes Tornado bug in a single request
Together AI has launched Alibaba's Qwen4 Max and demonstrated feeding all 344 files and 652,519 tokens of Tornado in a single request; the model produced a patch within 72 seconds that passed the regression test written by the maintainer for that bug, at a request cost of about 69 cents.
Industry news
Anthropic says Claude proved the Riemann Hypothesis; hours later, Stanford releases a fully human competing draft
Anthropic mathematician Levent Alpöge claims to have proven the Riemann Hypothesis and its generalized form GRH in collaboration with Claude, releasing a machine-verified proof in Lean split across 14 data packages.
AWS discloses: China-linked attacker used Cairn AI agent to autonomously discover zero-day vulnerability
Amazon Threat Intelligence disclosed that a China-linked threat actor used the open-source autonomous AI penetration-testing framework Cairn to have the agent autonomously discover and exploit a path traversal zero-day in an open-source medical imaging AI application, given only the goal "obtain shell access," achieving a root-privilege reverse shell in about 11 minutes.
US tech stocks post worst weekly performance since the pandemic as AI doubts mount
The Nasdaq 100 fell 5.6% on Friday and 10.4% for the week, its worst weekly performance since the pandemic, sitting 20% below its October high and entering bear market territory. Doubts about hundreds of billions of dollars in AI infrastructure spending, reports that Nvidia's accounting is under SEC scrutiny, and post-midterm policy uncertainty combined to trigger the selloff; Nvidia fell 14% on Friday and the US semiconductor index fell 10.7%. Anthropic fell 9.1% on its second day of trading.
Report: Chinese attackers deployed an AI agent that autonomously discovered and exploited a 0-day to compromise a university server
The Register reports that Chinese attackers deployed an AI agent that autonomously discovered and exploited a 0-day vulnerability to compromise a university server. The Hacker News entry contains only a link and title, with no further technical details.
Tips and opinions
Can Anthropic still put safety first after going public?
The Verge's Decoder podcast discusses Anthropic's situation after its Nasdaq listing; guest Hayden Field argues the company will need to significantly grow revenue over the next decade to support roughly $900 billion in compute spending.