Edited by the llms.blog desk.
The PHP2Shell embargo dispute, read against the rules. What CISA's disclosure program, CERT/CC guidance, ISO/IEC 29147 and 30111, and the EU Cyber Resilience Act say about ending an embargo early, and what they do not say about bugs found by AI.
A draft of Anthropic's IPO prospectus, reported by the WSJ, shows $16.3 billion in first-half 2026 revenue and a net loss of more than $80 billion, $72 billion of it accounting charges. Here is what the numbers mean, the risks it lists, and the IPO timeline.
A Tsinghua paper lets agents drop and reposition KV cache instead of summarizing history. Its AKV engine, built on SGLang, cut API-equivalent cost 27.7% and added 1.7 points on average across three open models and four agent benchmarks, with 1.51x faster P95 latency.
Mercury 3, Claude Haiku 5.5 and GPT-6 Luna compared on speed, API price, context and quality, using only figures from Inception, Anthropic, OpenAI and Artificial Analysis.
An agent reached a public chatbot through DNS and got little. OpenAI's monitor flagged it in under 12 minutes, but the run kept going for 2.5 hours. Why the pause on its most capable models is right, and what a restart should prove.
OpenAI, PerturbAI and Scripps Research say GPT-Rosalind agents, one per gene disruption, worked through a whole-brain mouse CRISPR atlas in 32 minutes. Experts preferred their reports 75% of the time. The leads on Tsc2 and DCTN1 are hypotheses that still need lab tests.
Meta's Q3 revenue rose 26% to $64.6 billion, but $34.9 billion in capex pushed free cash flow to negative $2.1 billion. It also announced nonvoting Class C stock that it says can help finance growth; Reuters says it may avoid diluting Zuckerberg's control.
VS Code 1.143 makes the Copilot harness the default for new agent sessions, adds a Compare Models preview that runs one task through Copilot, Claude and Codex harnesses with an optional judge, and lets you run MAI Code 1.1 Flash on a 96 GB GPU.
Google Cloud is adding GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna to Model Garden in private preview, with Codex support and billing against existing cloud commitments. Here is what is live, what is coming, and why both rivals want it.
fal and Black Forest Labs released FLUX 3 Max, a post-trained FLUX 3 video model that fal says makes a 10-second clip in 12.45 seconds. What it does, how solid the speed and quality claims are, and what it costs at $0.0425 per second until November 7.
Notion Agents now get a computer and web browser and can run tasks for up to 24 hours. Business and Enterprise seats get 1,500 Standard and 500 Select credits a month, extra credits cost $10 per 1,000, and existing workspaces switch on November 2.
Recursive's Flint is a 40B MoE coding model with 2.5B active parameters, released under Apache 2.0. Here is what it scores in Recursive's own tests, how fast it runs on a Mac or RTX Spark, the hardware you need, and what the Eureka Machine built.
Anthropic's Claude Duet, a proactive agent on Opus 5.5, opened to a US Pro and Max waitlist. Hacker News and X users praise the model and work-first pitch, object to background checks using plan limits, and question password access, VM transparency and lock-in.
Bria's Fibo 2.0 is an 8B model that edits and generates images from structured JSON prompts, trained on licensed data. Here is how its five modes work, Dev vs Edit Turbo, what Bria claims on benchmarks, and why the weights are non-commercial while hosted use costs $0.04 per image.
NVIDIA's DSX POD is a reference design for containerized Vera Rubin AI factories, built and tested by partners before delivery. Here is what is inside one container, how deployment changes, what it does not fix, and who ships it in early 2027.
OpenAI is bringing ChatGPT Ads to 26 more countries, for 89 in total, and says over 100,000 advertisers have used it. A new engineering post explains how an LLM orchestrator, kept apart from ChatGPT's answers, picks ads. Ads stay on Free and Go only.
Ten real AI tools for spreadsheets, grouped by job: Copilot in Excel, Gemini in Google Sheets, Claude for Excel, Shortcut, ChatGPT, Julius AI, GPTExcel, Numerous.ai, Rows and Better Analyst. What each does, its main limit, and prices only where the vendor's page states them.
Google started rolling out Personal Intelligence on Android Auto, letting Gemini use your Gmail, Calendar and Photos while you drive, and added Gemini to Google Maps navigation on CarPlay in the US. What each does, who is eligible, and the privacy settings.
Hugging Face CEO Clem Delangue says Netflix's Instadocs: AI Gone Wild went 'full doomer mode' on the July attack by OpenAI evaluation agents. Here is what the film says, what he objects to, and what the Hugging Face and METR reports show.
DeepSeek open-sourced DeepInfer, the Apache 2.0 inference engine behind DeepSeek-V4.1-Flash, for NVIDIA SM100 and Huawei Ascend 950. What it includes, what it leaves out, and what DeepSeek's claim of 155 trillion tokens a day and $38 per GPU-hour does and does not show.
