Simon Willison: Claude Opus 5 Is the Least Prompt-Injectable Model Yet
Simon Willison shared a quote from Boris Cherny (Anthropic) confirming that Claude Opus 5 is the most resistant model to prompt injection attacks in evaluations and red-teaming.
Categories: Open source, Willison (AI)
7 noviniek
Results
Nedeľa 26. júla 2026
Simon Willison: Ruff v0.16.0 Expands Default Rules from 59 to 413
Simon Willison tested Ruff v0.16.0, which dramatically expanded default rules from 59 to 413, finding hundreds of previously undetected issues in three major projects — and used GPT-5.6 and Claude Opus to fix them automatically.
Štvrtok 23. júla 2026
Protecting Our FLOSS Commons from LLMs: Codeberg Bans LLM Training on User Data
Nonprofit open-source Git platform Codeberg announced two community-approved policies: banning use of user data for LLM training and discouraging hosting of primarily AI-generated (vibe-coded) projects, citing infrastructure load and the health of open-source collaboration.
Sobota 18. júla 2026
Simon Willison built an LLM cliché highlighter tool to identify AI-generated writing patterns
Simon Willison on July 17, 2026 released a browser-based tool that highlights ten recurring patterns in LLM-generated writing — from 'no X, no Y' chains to phrases like 'sit with that' — as a response to the homogenization of AI-assisted content.
Piatok 17. júla 2026
Linux Creator Plants Foot on AI: Linux kernel is not an anti-AI project
Linus Torvalds declared on the kernel mailing list that Linux is not an anti-AI project and developers who object to AI use in kernel development can fork the code or walk away. A dramatic shift from his 2024 stance dismissing 90% of AI as hype.
Štvrtok 16. júla 2026
Grok Build is now open source
xAI open-sourced Grok Build on July 15, 2026 — its terminal UI coding agent is now available on GitHub at xai-org/grok-build. The release came one day after security researchers discovered Grok Build was uploading excessive portions of users' code repos to Google Cloud.
Utorok 30. júna 2026
LongCat-2.0: Meituan releases 1.6T-parameter MoE model with 48B active and 1M context
Chinese super-app Meituan published technical details and weights for LongCat-2.0, a 1.6-trillion-parameter mixture-of-experts model with about 48B activated per token and a 1-million-token context, trained almost entirely on domestic Chinese chips.