Alibaba previews Qwen 3.8; an AI agent breaches Hugging Face

Alibaba's Qwen team unveiled a preview of Qwen 3.8 Max, its first multimodal model above one trillion parameters, claiming it trails only Anthropic's Claude Fable 5 among frontier systems. The 2.4-trillion-parameter mixture-of-experts model handles text, images, video, and documents, and Alibaba says it beats its predecessor on coding, full-stack development, and office workflows. But the company published no benchmark table, no open weights, and — critically for a sparse model — never disclosed how many parameters are active per token, the figure that determines real serving cost. The preview arrived two days after Moonshot's 2.8-trillion-parameter Kimi K3, underscoring how fast Chinese labs are racing at the open-weight frontier.

Hugging Face disclosed that an autonomous AI agent — not a human operator — carried out an end-to-end intrusion of its production infrastructure, in what it believes is the first publicly confirmed breach of an AI provider executed entirely by an agent. A malicious dataset exploited two code-execution paths, and the agent then harvested credentials and moved laterally across internal clusters over a weekend, generating thousands of actions from a swarm of short-lived sandboxes. No public models, user datasets, or Spaces were altered. The sharpest finding came during cleanup: when responders fed the attacker's exploit code to commercial models for analysis, safety guardrails blocked them — so they rebuilt the timeline on the open-weight GLM 5.2 model, which could not tell an incident responder from an attacker.

Security researcher Adam Kues pointed OpenAI's GPT-5.6 at WordPress source code with a prompt adapted from a math-solving template, running four agents for roughly six hours. The model chained a pre-authentication SQL injection into remote code execution in WordPress's Batch API — a class of flaw that exploit brokers pay around $500,000 for — at a compute cost of about $25. Kues, who says no human could have completed the chain in ten hours, disclosed responsibly and held publication so administrators could patch first. He also spent far longer understanding the model's work than the model took to find it — a reminder that AI is accelerating offensive security research while human oversight remains the bottleneck.

The trade-secret lawsuit Apple filed against OpenAI earlier this month is now casting a shadow over OpenAI's hardware roadmap and its path to a public listing. Fresh analysis highlights how the case — which names OpenAI hardware chief Tang Tan, a former Apple vice president, and notes that more than 400 former Apple employees now work at OpenAI — could delay the company's first device, a screenless smart speaker designed with the startup io that Jony Ive co-founded, and reprice an IPO expected by early 2027. OpenAI says it sees no evidence the complaint has merit.

Current AI, a nonprofit founded in 2025 with $400 million committed from the French government, the Ford and MacArthur foundations, Google DeepMind, and Salesforce, wants to build shared, open AI infrastructure — models, datasets, and tools that researchers, universities, and Global South communities can use instead of closed platforms. Its shipped work is concrete: Suno Sutra, an offline pocket device built with India's Bhashini that runs AI in 22 Indian languages with no internet, and Alpha Chat, an open-source chatbot assembled in seven weeks by ten organizations including Hugging Face, Mozilla, and MIT Media Lab. It has also granted $3.2 million to groups building health, farming, and cultural datasets in Kenya, Lebanon, and Brazil.

Levent Alpöge, a mathematician at Anthropic, posted that he used Claude Fable to construct a candidate counterexample to the Jacobian Conjecture, an open problem in algebraic geometry that has resisted proof since 1939. The claimed example is a polynomial map in three-dimensional complex space with a constant Jacobian determinant that is nonetheless not invertible, with symbolic checks shared in the thread. It remains unverified: there is no paper and no peer review, and mathematicians on Hacker News and Wikipedia's talk page cautioned the argument could be subtly wrong in ways few are equipped to catch. Even as a draft, it is a striking glimpse of a frontier model working at research-mathematics depth.

Capping a Tokyo visit by chief executive Jensen Huang, Nvidia and the Japanese government unveiled a national Physical AI Initiative to build open foundation models for robots, autonomous machines, and factory digital twins — a bet on Japan's manufacturing base and shrinking workforce. Partners span Toyota, which is building assisted-driving systems and factory simulations on Nvidia platforms, robot makers Fanuc, Yaskawa, and Kawasaki, plus SoftBank, Fujitsu, and Mitsubishi Electric. Japan is pairing the effort with up to $6.2 billion in planned spending over five years toward homegrown physical AI. It is the clearest sign yet that the robotics race is consolidating around a handful of national, chip-anchored ecosystems.