🚨 Meta wants to sell $95B of stock to pay for AI

The FT reports Meta plans an equity offering to fund its AI buildout. Stock, not bonds. Existing shareholders eat the dilution.

Even a company that prints cash by the billions is reaching for outside money. Meta hasn't confirmed the size or timing.

🧬 Anthropic now has in-house drug programs, plus six pharma and biotech partners

Anthropic published early results with BMS, Generate Biomedicines, Enveda and three others. At Generate, Claude Code agents proposed 186 CAR T designs for lab testing, and one came out about 70% shorter with roughly 10x the surface receptors. Anthropic says the work is mostly preclinical, so almost none of it has reached humans yet.

🚨 Google's LLM bug hunter beat its best JS fuzzer, but only 45 to 37

Project Zero compared BigSleep and Fuzzilli on V8 over six months. BigSleep found 45 actionable vulnerabilities, Fuzzilli 37, and only about 25% of the bugs overlapped.

Re-run on old Fuzzilli finds, BigSleep caught roughly 60%, or 70% with tweaked prompts. It did worse on JIT compiler bugs.

🚨 Delaware wants AI agents to legally run companies, with no human in charge

Ars Technica reports on a proposal that would let autonomous systems operate entire businesses in the state where most big US companies are incorporated. It's still only a proposal, and it doesn't say who gets sued when an agent breaks a contract.

🤖 Cursor ships the framework it uses internally to build its own agents

Cursor released the Agent SDK. You define agents in code and deploy them to Cursor's infrastructure. It's the same stack behind Bugbot, Security Reviewer and Rollouts.

• Agents react to events like pull requests on GitHub, GitLab or Bitbucket
• One command deploys: agent-sdk deploy
• Example job: assess PR risk, route it to the right team, auto-approve what meets policy

🚨 Google's Gemini 4 Carbon takes #1 on the Artificial Analysis index, and you can't use it yet

Google announced Carbon two weeks after Argon, pitched on cybersecurity and long-horizon work:

• Scores 59 on Artificial Analysis v5.0
• SEC-bench Pro 88.2% (finding vulnerabilities)
• ExploitGym 54.8%
• CWE-bench 71.0% (patching)

Access starts with a small group of cyber defenders. Everyone else waits, and the price is $2 per million input tokens, same as Argon.

🤖 DeepSeek-V4.1-Flash is bigger than V4-Flash and needs about a quarter of the KV cache

DeepSeek's paper describes a multimodal MoE with a 1M-token context:

• 552B backbone params, using a Causal Encoder-Decoder design
• 8B active per token in prefill, 16B in decode
• Roughly 1/4 the runtime KV cache and 1/8 the persistent KV cache of V4-Flash
• 437x smaller per-token cache than DeepSeek-V1

It's cheap where agents hurt most, which is input-heavy prompts.

🚨 The DOJ is asking questions about Nvidia buying Hugging Face

Reuters reports that antitrust staff have started inquiring about the deal, citing anonymous sources. The biggest AI chip company would own the place where open models live.

It's an inquiry, not a formal challenge. No timeline has been reported for the deal.

🤖 Nvidia's new open ASR model covers 40 languages, but it's batch-only

Nvidia released Nemotron 3.5 Transcribe, built for recorded audio, not live streaming. It also has experimental near-field speaker transcription, which no other open speech LLM offers yet.

Early testers say it's slower, heavier and less accurate than Parakeet. Some prefer Qwen ASR too.

🚨 A startup says it got into Anthropic's systems in 12 hours

Cathedral, founded by ex-DOGE staffers, claims it reached employee details and internal messages. It was reportedly done through a bug bounty program, so it may have been authorized.

A senior Trump administration official allegedly said that if a bunch of kids can do this, the Chinese aren't far behind. Anthropic's side and the details are still unconfirmed.

🚨 Claude's new smart reports let Enterprise admins see which sessions went "off-topic"

Anthropic's beta feature, covered by Mashable, reads a sample of one team's Claude transcripts from the last 28 days. It reports the work done, the cost, and the count and price of sessions that were off-topic or produced nothing usable. Anthropic says it's for adoption planning, not judging individuals.

But it's a boss-readable log of what you asked the AI.

🇪🇺 The first AI Act harmonised standard is out, and it covers quality management only

The Commission published EN 18286:2026 in the Official Journal. Providers of high-risk systems who follow it get a presumption of conformity with Article 17 (quality management), and it's built to sit on top of ISO/IEC 42001, ISO 9001 or ISO 13485 instead of replacing them.

🚨 Gemini 4 Carbon takes #1 on the Artificial Analysis Intelligence Index v5

It's the first independent ranking since launch, and Artificial Analysis puts it ahead of GPT-6.1 Terra (Max), which scored 48 on the same index.

But Carbon is still limited to a small group of cyber defenders, so most developers can't touch the model that just topped the chart.

🚨 Mistral Large 4 goes open weight, and it's a 1T-parameter MoE

Mistral released the weights with 52B active per token, plus an NVFP4 quantized variant on Hugging Face. The license is a modified MIT that requires a commercial license if you serve it as a business, so it isn't OSI-approved open source.

🇰🇷 FuriosaAI quadruples its valuation to $2.9B on a $620M raise

The South Korean chip startup is one of the few non-US names taking money to go after Nvidia directly. Investors are betting its inference chips can win real data center slots, which no startup has done at scale yet.

$2.9B is still about 0.06% of Nvidia's market cap.

🚨 A defense contractor plugged his personal AI agent into a secure Pentagon system

Wired reports that the agent ran with the contractor's own credentials. It reached a system that was never meant to touch a personal tool.

That's shadow AI inside a government network. Nobody sandboxed the agent, and nobody scoped its permissions.

🇨🇦 Canada is putting up to C$1.6B into Cohere

Ottawa is backing the Toronto-based lab, CBC reports. Cohere is one of the few non-US labs training its own models, and now it has a government behind it.

It's an "up to" figure, so the final amount depends on the deal terms.

🪦 Google is killing Jules, its async coding agent, on December 31, 2026

Sessions go read-only after November 11. Google says the lessons are feeding "Google Antigravity", its agent-first platform.

Jules launched as an experiment in handing your backlog to an agent. Now it's one more coding tool folded into a bigger one.

🚨 Open-weight models can now turn a patch into an exploit before users update

Epoch AI shows that open-weight models like GLM 5.3 can exploit the patch gap. That's the window between a fix landing publicly and users getting the release.

Mozilla says it's working on a response. It won't ship renderer fixes and sandbox-escape fixes in the same gap.

🚨 A trojanized Claude Code extension is planting an SSH backdoor on developer machines

OpenSourceMalware details a fake extension that gives attackers persistent SSH access to whoever installs it. If you've added any Claude Code extensions lately, audit them and check your authorized_keys for keys you don't recognize, because uninstalling the extension alone won't close a persistent backdoor.