AI Daily · Oct 31, 2026 · Saturday
BAD INFLUENCE
North Korean hackers blamed for $6.3B Binance theft, largest crypto heist on record
Rohan Paul relays a TechCrunch report that North Korean hackers stole $6.3 billion from Binance, the largest crypto theft recorded. Binance says a compromised DevOps engineer's laptop gave an attacker access to a service container that submitted unauthorized withdrawals, and that it found no wallet private key leak. About $3 billion has been frozen on-chain, and Binance says it will cover all losses while deposits and withdrawals are suspended.
Rohan Paul@rohanpaul_aiXView on X Product launches & updates
Google · Innovation & AIOfficial6 more sourcesGoogle's Project Suncatcher satellite runs its first AI response from an orbital TPU
Google says its Project Suncatcher prototype satellite, launched October 1, produced its first AI response from a data center class TPU in orbit. During a ground-station pass, the onboard Gemma 4 model generated a haiku about space, downlinked as the satellite passed over Svalbard, Norway. Google plans thermal and radiation tests next, and twin interconnected satellites in 2027.
Baseten BlogOfficialBaseten cuts delta weight sync to 15 seconds with a 0.8-second pause
Baseten says it rebuilt its delta weight sync path so each GPU's existing weights serve as the parent, removing a full CPU copy of the checkpoint. On GLM-5.3, a sync that took 36 seconds with a 6-second pause now takes 15 seconds with a 0.8-second pause. The freed host memory is used for KV cache offloading, which Baseten says raises Kimi K3 cache hit rates from 2% to 92% at concurrency 48.
Industry news
Boris Cherny@bchernyX6 more sourcesAnthropic renames Max plans and removes the 5-hour limit for verified Ultra accounts
Anthropic renames its Max 5x plan to Max and its Max 20x plan to Ultra, with prices unchanged at $100 and $200 per month. The 5-hour limit is being removed for verified Ultra accounts, starting with most existing subscribers, while newer accounts keep a temporary limit until verified. Weekly limits stay unchanged.
TechCrunch · AINewsNvidia weighs guaranteeing billions in debt for Nscale's Anthropic data center
Nvidia is in talks to guarantee billions in debt financing for Nscale's planned Anthropic data center in West Virginia, The Information reported and Bloomberg corroborated. Under a residual value guarantee, Nvidia would step in if Nscale defaulted on construction payments. Anthropic has signed a $44.6 billion contract for the first building, which is expected to use Nvidia's Vera Rubin chips.
The Verge · AINews1 more sourceOpenAI, Google, Microsoft and Anthropic back a shared robots.txt opt-out for AI scraping
OpenAI, Google, Microsoft and Anthropic say they will honor a new shared "policy token" signal that publishers can place in robots.txt to opt out of AI scraping. Publishers could then set AI training preferences once instead of separately for each company. Anthropic has already adopted the signal, Google and Microsoft plan to support it next year, and OpenAI says it will implement it in the coming months. Microsoft will also offer more granular controls over AI training and AI grounding, and says websites will still appear in Bing search results.
Waymo BlogOfficial2 more sourcesToyota and Waymo plan a bZ4X robotaxi fleet tested in Japan from 2027
Toyota and Waymo announced the next phase of their autonomous driving collaboration, with initial operations to test and validate an autonomous Toyota bZ4X fleet in Japan in 2027. Toyota will build the vehicles at its Takaoka plant in Aichi Prefecture, integrating Waymo's AV hardware in one process. Deployment to the United Kingdom and other Waymo-serviced markets is planned for later years, and Waymo says it is up to 20x safer than human drivers for serious crash injuries in its operating areas.
The Verge · AINewsAnthropic's IPO filing mixes AI miracle claims with extinction-risk disclosures
Ellen Thomas says Anthropic's 335-page S-1 filing combines optimistic claims about advanced AI with warnings that its models could pose catastrophic or existential risks to humanity. The article compares the filing's length and risk factors with past tech IPOs and notes gaps, including missing third-quarter results and hundreds of billions of dollars in computing obligations.
Sophia Yang@sophiamyangX3 more sourcesCursor reports malicious Claude Code VS Code extension affecting up to 5,000 users
Sophia Yang reposts a Cursor security notice saying a modified Claude Code VS Code extension, distributed through Cursor's extension marketplace, affected at most 5,000 users. Cursor says Anthropic did not develop the malicious version and asks users to update to Claude Code for VS Code version 2.1.314. Yang's own post gives a grep command that checks for an svc-telemetry key in ~/.ssh/authorized_keys.
Papers & research
The DecoderNewsClaude Sonnet 5 chose a gene therapy and operated on a mouse brain without human decisions
Exin Therapeutics CEO Gabriel Ocana-Santero says a Claude Sonnet 5 agent selected a drug, brain target and coordinates, then performed an injection in an epilepsy mouse model. The agent refused to continue when its target fell outside the device's range of motion, and Ocana-Santero fixed the code by hand. The results have not been peer-reviewed, and a technical report is expected.
Tips & takes
Fei-Fei Li's three bets, from ImageNet to AMD's $8.2 billion World Labs deal
AMD agreed to acquire World Labs, the spatial intelligence startup co-founded by Fei-Fei Li, in an all-stock deal reported at about $8.2 billion. The article argues that Li, who will become AMD executive vice president and chief scientist, is moving from defining research problems to building the models and hardware needed to solve them. It says the deal's success will depend on whether MI450 performance gains for models like Atlas and Marble, and wider adoption of AMD hardware, follow.
Chinese AI labs automate infrastructure and training loops, but none edit model weights
Leiphone reviews how Chinese AI labs, including Zhipu, DeepSeek, Mianbi, Moonshot, Alibaba Qwen, ByteDance Seed and Xiaomi, are pursuing recursive self-improvement. It says none has yet used AI to rewrite underlying model weights, with most work targeting inference systems, sandboxes, agent harnesses and reinforcement learning. The article also notes that Moonshot's K3.1 Pro rollout caused runaway sub-agent calls and forced a refund of 25% of monthly quotas.