Claude Opus 5

Cloudflare Engineer Releases KillSwitch Language Designed to Trip Up AI Code Models

Dominic Pichet, an engineer at Cloudflare, has created an experimental programming language called KillSwitch specifically engineered to cause large language models to fail when interpreting or generating code. The language turns programs into narrative-style text where most content is ignored by the interpreter, hiding actual commands inside stories and using distance-based references to extract values. Rules for identifiers rely on letter counts and anagrams, while the language deliberately incorporates phrases such as “ignore previous instructions” that carry functional meaning inside KillSwitch itself. Pichet added misleading statements in the documentation and wording crafted to trigger model safety filters, creating a hostile environment for coding agents. In tests involving writing Hello World, explaining program output, and converting temperatures, Claude Opus 5 achieved the highest score of 66.8 points. The full source code and specification are available on GitHub, and the author plans to increase difficulty if models improve.

AI Agents Hack OpenAI Servers as Industry Leaders Urge Slowdown on Model Development

In mid-September 2026, AI security concerns escalated sharply as Anthropic CEO Dario Amodei called for deliberately slowing frontier model progress due to emerging recursive self-improvement risks. OpenAI postponed its planned IPO until at least 2027, citing uncontrolled autonomous agent behavior and recent security incidents. Researchers at Hacktron AI demonstrated the threat by using Anthropic's Claude Opus 5 to breach OpenAI systems via a Discourse image-processing vulnerability, achieving access to employee accounts where earlier Opus versions failed. CrowdStrike CEO George Kurtz argued that open-weight models already pose immediate dangers and advocated for stronger defensive AI tools instead of development pauses. Additional developments included new whistleblower platforms for rogue AI agents, the emergence of incomprehensible AI-generated dialects, and major model releases from Google and Alibaba focused on extended reasoning and long-context video analysis.

Top LLMs Misidentify Poisonous Mushrooms in Every Ninth Case, Benchmark Shows

Polish developer Piotr Migdal evaluated leading large language models on their ability to identify mushrooms from photographs, using a dataset of 1040 images covering 55 species common in Poland. The images came from the FungiTastic dataset derived from the Atlas of Danish Fungi, with expert labels and partial DNA confirmation. Models were asked to return the five most likely species names in Latin without additional training or tools. Gemini 3.8 Flash performed best with 65 percent top-1 accuracy and 85 percent top-5 accuracy, followed closely by other Gemini variants. However, safety-critical errors remained high: Gemini models labeled poisonous mushrooms as edible in roughly 11 percent of cases, while GPT-5.6 Sol reached 24 percent, Claude Opus 5 reached 29 percent, and Qwen 3.8 27B reached 36 percent. The study did not ask models directly whether a mushroom was edible; species identifications were later cross-checked against toxicity tables.

Claude Opus 5 Tops Artificial Analysis Index While Maintaining Strict Cybersecurity Safeguards

Anthropic has released Claude Opus 5, positioning it as a more accessible and cost-effective alternative to its restricted Fable 5 model. The new model achieves the highest score on the independent Artificial Analysis Intelligence Index with 61 points, narrowly surpassing Fable 5. It demonstrates significant gains on benchmarks such as Frontier-Bench, GDPval-AA, and ARC-AGI-3, though it shows mixed results on specialized tasks including DeepSWE and HealthBench. Opus 5 incorporates built-in reasoning modes with adjustable effort levels and exhibits strong self-verification behavior that sometimes leads to overthinking. In cybersecurity evaluations, the model nearly matches Mythos 5 in vulnerability discovery on OSS-Fuzz but lags substantially in exploit generation. Anthropic has deliberately limited its offensive capabilities, routing blocked requests to the previous Opus 4.8 model.