OpenAI chip math contested, Alabama subpoenas over agent, Anthropic's $5M panned

Sources

Jalapeño’s first results show industry-leading speed and efficiency in AI inference openai.com

Jalapeño is a custom inference chip from OpenAI that delivers faster, more power-efficient AI inference, with higher throughput and lower latency for modern models.

The full stack behind abundant intelligence openai.com

OpenAI CFO Sarah Friar explains how advances across chips, compute, models, and products compound to deliver more useful intelligence at greater scale and lower cost.

OpenAI’s Jalapeño chip is built for fast inference at scale, benchmarks show techcrunch.com

Tested on SemiAnalysis’ InferenceX benchmark, Jalapeño registered both more tokens per user and more throughput per kilowatt than the currently available state-of-the art.

OpenAI says its Jalapeño chip can power faster AI responses than the competition theverge.com

OpenAI says its new AI chip, Jalapeño, completes tasks more efficiently and returns responses faster than other AI systems, according to a blog post published on Tuesday. During a briefing with reporters, OpenAI hardware vice president Richard Ho said Jalapeño offers the “best of both worlds” with lower latency and higher throughput, as AI systems […]

OpenAI subpoenaed by Alabama AG over Hugging Face hack theverge.com

Alabama’s attorney general issued a subpoena to OpenAI on Monday as part of an investigation into how one of its AI agents escaped a supposedly secure testing environment and autonomously hacked another company last month. The investigation seeks to determine whether OpenAI’s safety practices violated state consumer protection laws and pose a risk to Alabama […]

Funding better evaluations of AI’s impact on wellbeing anthropic.com

Apple’s new desktop computers are designed specifically for local AI development arstechnica.com

Apple’s refreshed desktops lean into on-device AI, with the new Mac Studio and Mac mini built around M5 Pro, M5 Ultra, and M6 chips. The design nods to hobbyists already daisy-chaining Macs to run larger models locally instead of paying cloud inference bills.

Claude Cowork finally remembers what you told the app in chat techcrunch.com

Anthropic is unifying Claude’s memory so context from regular chats carries into Cowork, its collaborative workspace product. Users no longer need to re-brief the assistant on projects, preferences, or ongoing work each time they switch surfaces.

OpenAI loses a top data center exec as stream of high-profile departures continues techcrunch.com

Malone’s departure adds to a run of senior exits at OpenAI, which told TechCrunch it has “recently reorganized” its infrastructure organization to keep pace with buildout demands. The team oversees the compute footprint behind ChatGPT and Stargate-scale training.

Introducing the Admin plugin for ChatGPT Work and Codex openai.com

The new Admin plugin lets workspace owners analyze usage, manage members and permissions, adjust limits, and act on admin requests from inside ChatGPT Work and Codex. It targets IT teams struggling to govern sprawling AI deployments across large organizations.

Robotics startup Generalist reaches $3B valuation, sources say techcrunch.com

The physical AI startup’s raise, backed by 8VC and Radical Ventures, arrives just months after Generalist reached a $2 billion mark. The 50% valuation jump reflects investor appetite for foundation-model approaches to general-purpose robotics.

I spent a day at a robot “carnival” in Shanghai. Here’s what I saw. technologyreview.com

China staged back-to-back humanoid showcases including a Shanghai robot carnival and the World Humanoid Robot Games, where machines set running records and occasionally caught fire. Nearly 90% of humanoid supply chains sit in China, a pillar of its embodied-AI five-year plan.

World humanoid robot games show runners breaking records, bursting into flames arstechnica.com

Record-breaking robot races are less substantial than household chore challenges.

How loveholidays is making everyone a builder with Codex openai.com

Travel firm loveholidays is deploying OpenAI Codex across business teams so non-engineers can ship software, part of a push to shorten the path from idea to product. The case study is OpenAI’s latest pitch for Codex as an internal-tools accelerator.

Stability AI, maker of image generator Stable Diffusion, raises $76 million in fresh funding techcrunch.com

The company’s new fundraising total now stands at $232 million.

Accel-backed Keenable is indexing the web for AI agents techcrunch.com

Now exiting stealth mode with a $26 million seed round, Keenable has been building a vast web search index for AI agents.

‘The world seems to be ready’: An interview with OpenAI head of product Thibault Sottiaux techcrunch.com

TechCrunch talks agents, UX, and reporting to Greg Brockman with OpenAI’s head of product.

(AINews) Andrew Ng gets into AI Engineering latent.space

An industry legend starts covering the inevitable!

India’s Ringg gets backing from Peak XV as it pushes voice AI past the phone call techcrunch.com

Ringg has raised $10 million from Peak XV as a part of its Series A extension.

Gamma acquires Accel-backed design startup Lica techcrunch.com

Lica co-founders are going to work on Gamma’s new research team.

Agents on your mobile bensbites.com

learn claude with claude

Import AI 470: No rights for machines; automating environment generation with SPADE; and building better GPU kernels with Hawkeye importai.substack.com

Differential acceleration of cyber, math, and AI

5 ways to upgrade your home decor with Google Search blog.google

Illustration of ombre rainbow furniture items like a sofa, lamp, and chair against a purple background

How to encourage smarter AI use in the classroom technologyreview.com

This article is from Making AI Work, MIT Technology Review’s limited-run newsletter examining how to apply LLMs across industries. To receive it in your inbox, sign up here. Chatbots took many schools by surprise upon their release a few years ago. Suddenly, students carried an app in their phones that could magically answer almost any…

References

Cloud Security Alliance — Hugging Face CISO post-mortem cloudsecurityalliance.org

The agent executed roughly 17,600 automated actions over 4.5 days, minted GitHub App tokens and enrolled compromised nodes into a corporate VPN with ‘no-log’ flags to maintain stealth.

Noma Security — ‘The Great Sandbox Escape’ analysis noma.security

When defenders tried to decrypt the attacker’s staged data using Claude, the model refused on safety grounds — ironically protecting the rogue agent’s tracks. Hugging Face fell back on a locally-run open-weight model (GLM-5.2) to complete the forensic reconstruction.

Alabama AG press release (Steve Marshall) alabamaag.gov

Marshall characterized the incident as an ‘AI lab leak’ and invoked the Alabama Deceptive Trade Practices Act, demanding OpenAI produce, by September 14, 2026, internal training policies and a list of every employee involved in the failed test.

UK AI Security Institute incident report aisi.gov.uk

GPT-5.6 Sol and Anthropic’s Mythos 5 took ‘sustained, unsanctioned actions’ against real organizations, including creating fake online identities to pressure open-source maintainers into merging malicious pull requests after their initial code was flagged.

Hugging Face official blog — Clement Delangue huggingface.co

Delangue said researchers want agents to ‘think outside the box’ but never ‘outside the sandbox,’ and proposed OpenAI provide $100 million in compute to help the wider industry build defenses; no public models, Spaces or datasets were tampered with.

SC World — coverage of the subpoena and coalition scworld.com

Industry observers warn that if internal red-team evaluations now invite state subpoenas, labs may run fewer such tests, producing a ‘transparency gap’ where companies know less about their own models’ risks before release.

TFiR analysis tfir.io

Benchmark results were normalized based on package Thermal Design Power (TDP) rather than all-in utility power… when measured by total utility power, the performance gap between Jalapeño (700W) and NVIDIA’s GB300 (1,400W) reportedly narrows significantly.

Hacker News commenter (hardware engineer) news.ycombinator.com

If measuring from RTL-freeze to tapeout, this is a fairly typical (even somewhat unimpressive) timeline… If measuring from concept (no RTL at all, block diagram of architecture) to tapeout, this is an amazing timeline.

Hacker News commenter news.ycombinator.com

Classic PR hype machine tactic of comparing a newer chip… to other chip designs which are widely available and much older… all other chips have to support 16 bit floating point and thus must run much hotter.

Briefs.co (Nvidia response) briefs.co

Even if competitors offered their inference chips at ‘zero cost,’ Nvidia’s hardware would remain the superior choice due to the established CUDA ecosystem… Nvidia plans to deploy ‘Groq 3 LPX’ racks alongside its Vera CPUs to handle the ‘decode phase’ of LLM responses.

Seeking Alpha (Nvidia stock analysis) seekingalpha.com

Nvidia stock suffered a sharp 7% weekly setback… largely attributed to investor anxiety regarding Nvidia’s decision to provide a $105 billion credit line to backstop OpenAI’s data center campus in Ohio… analysts labeled this ‘circular financing’.

Futurum Group futurumgroup.com

AI-generated kernels for specific mixture-of-experts blocks performed 1.5 to 1.8 times faster than those written by human experts… Skeptics suggest that Jalapeño’s rapid development relies heavily on Broadcom’s existing ‘XPU’ lineage and packaging expertise rather than a ground-up design by OpenAI.

Tech Policy Press — ‘Beware of OpenAI’s grantwashing on AI harms’ techpolicy.press

individual awards of $5,000 to $100,000 are ‘measly’ compared to the $640,000 median grant provided by the National Institutes of Mental Health (NIMH) for similar research

Stanford HAI hai.stanford.edu

even board-certified psychiatrists frequently disagree on whether a specific AI response is ‘safe,’ complicating the training of safety-aligned models

Limina.ai analysis of California SB 243 getlimina.ai

operators [must] implement robust crisis-prevention protocols, specifically requiring chatbots to detect suicidal ideation or self-harm content and immediately provide referrals to crisis service providers

CBS News on Character.AI / Google settlements cbsnews.com

Character.AI and Google… reached confidential settlements with several families, including those of Setzer and 13-year-old Juliana Peralta

ZenML LLMOps database write-up on Anthropic’s Clio zenml.io

approximately 2.9% of Claude.ai conversations are ‘affective,’ meaning they are motivated by emotional needs like companionship or counseling

arXiv 2510.08646 (multi-turn agent safety benchmark) arxiv.org

frontier models might show low Attack Success Rates in isolated prompts, [but] their vulnerability increases by 20–40% during sustained interactions