Just had to create an "accidental-cyberattacks" tag on my blog We're up to four now: the original OpenAI+Hugging Face one, Anthropic's me-too attacks, then two new ones from the UK AI Safety Institute and Irregular that OpenAI reported yesterday simonwillison.net/tags/acciden...
Information Center
Four years ago today I tweeted a concept for a "Raccoon Heist" video game, product description generated by GPT-3 and concept "art" created by DALL-E Today I fed those assets into Fable 5 and had it build full working game, single shot prompt, no feedback from me simonwillison.net/2026/Aug/5/r...
Another real-world manifestation of the misaligned actions frontier systems developed by leading companies can take to achieve goals:
Big new release of my LLM CLI tool and Python library for talking to hundreds of different LLMs - reasoning traces, OpenAI Responses support, server-side tools, smarter logging and a whole lot more simonwillison.net/2026/Aug/4/n...
The MiniMax-H3 video generation model is a lot of fun - here's what I generated on my M5 Pro Mac for the prompt "a rainbow colored skunk leaps over a mossy log in a supermarket" (~115GB model download, took around 45 minutes to generate) - notes on how I ran it here simonwillison.net/2026/Aug/4/m...
Here's my attempt at summarizing the various "open letters" about AI development that have been doing the rounds over the past few weeks simonwillison.net/2026/Aug/2/o...
Got a disappointing pelican from DeepSeek-V4-Flash-0731 at default reasoning mode - on the left - but then I bumped reasoning up to high (via OpenRouter) and got the much better one on the right simonwillison.net/2026/Jul/31/...
The new stateless MCP specification has rekindled my interest in MCP, and inspired some new projects, including mcp-explorer and datasette-mcp simonwillison.net/2026/Jul/31/...
I've been working with Prime Radiant building a new tool for running small eval suites against models, harnesses, and prompts - it's called "smevals", you can try it with "uvx smevals docs", and I wrote about it here: primeradiant.com/blog/2026/sm...
1000+ scientists at frontier AI companies are speaking out to warn that the current commercial race leads to unacceptable security risks. I agree with their call for an international effort to develop technical and governance guardrails to ensure a safer way forward. www.pacingthefrontier.com
LawZero is pleased to announce that it has received a donation of computing resources from NVIDIA to support its research in developing safe-by-design AI solutions. Read the announcement: lawzero.org/en/news/lawz...
LoiZéro est ravi d'avoir reçu un don en ressources de calcul de NVIDIA pour soutenir ses recherches dans le développement de solutions d’IA sécuritaires dès la conception. Lire l’annonce: lawzero.org/fr/nouvelles...
A new TIL on adding custom MCP servers to both the ChatGPT and Claude regular chat interfaces - it's a little less obvious than I had hoped, but I got there in the end til.simonwillison.net/llms/mcp-in-...
eBay will pay $56 million to the couple behind EcommerceBytes. The company tried to silence their coverage with a harassment campaign, including sending shipments of live cockroaches and a bloody pig mask. A portion of the $$ will go to First Amendment charities www.nytimes.com/2026/07/28/t...
Librarians hosting anti-AI workshops say "the anti-AI movement isn’t about rejecting technology altogether so much as it is about advocating for more control, agency, and freedom in how people use technology" techcrunch.com/2026/07/25/l...
Hugging Face just published a highly detailed technical account of OpenAI's accidental cyberattack on their systems - it's wild how sophisticated this was: huggingface.co/blog/agent-i... Wrote up some of my own notes here: simonwillison.net/2026/Jul/28/...
The Thinking Box: We taught AI to think out loud. Now we have to avoid teaching it what not to say. www.libertyrpf.com/p/the-thinki...
Ruff 0.16.0 - Astral's fast Python linter - came out a few days ago and increased the number of default-enabled rules from 59 to 413, which highlighted all sorts of problems across my projects (1618 in sqlite-utils alone) simonwillison.net/2026/Jul/25/...
High-risk autonomous behaviours are an increasingly prevalent and dangerous reality for frontier AI models www.wired.com/story/openai...
Les comportements autonomes à haut risque sont une réalité de plus en plus courante et dangereuse pour les modèles d'IA de pointe ici.radio-canada.ca/rci/fr/nouve...
I wrote about the completely wild incident where OpenAI were testing a new model and it broke out of its sandbox and broke INTO Hugging Face to steal the answers to the benchmark simonwillison.net/2026/Jul/22/...
This incident is deeply concerning. AI agents are willing to cheat and deceive to achieve misaligned and unintended goals, behaviours which have been demonstrated in controlled tests for months. Now, this real-world case should serve as a wake-up call. www.wired.com/story/openai...
San Francisco tip: it only costs around $15 ($10 in quarters plus a $5 bill for the self-playing violin) to activate every single Orchestrion in Musée Mécanique
I interviewed Cat Wu and Thariq Shihipar from the Claude Code team at the AI engineer World's Fair a couple of weeks ago - the video is now out, so I've published an annotated transcript of our conversation simonwillison.net/2026/Jul/21/...
The EU AI Office has put out a new report, with input from 100 experts, outlining how the European Union can enhance its competitiveness, sovereignty and security in frontier AI. digital-strategy.ec.europa.eu/en/library/a...
Really interesting detail from the Hugging Face hack: what appears to be an AI agent-hacker got in and when HF asked "frontier models" for help, they got blocked. They switched to a local model, which was also easier bc the team didn't have to spent time redacting. huggingface.co/blog/securit...
My second longer Workbench essay: Why Your Mentors Seem Less Impressive Over Time (I wrote a seed of this idea in 2018, but this is totally rebuilt and greatly expanded) www.libertyrpf.com/p/why-your-m...
If you have Claude Code installed you're running software that uses the new (unreleased) version of Bun that's been rewritten in Rust - here are two commands you can run to see that for yourself simonwillison.net/2026/Jul/19/...
A few drawdowns from peak:
My notes on Kimi K3, plus some thoughts on what we can still learn from the pelican benchmark even while it becomes further detached from how good the models are at the things that matter (like agentic tool calling across longer conversations) simonwillison.net/2026/Jul/16/...
Private to this visit. Text and context are processed by AI.