- Anthropic Releases Claude Opus 4.6, Claiming Major Improvements in High Severity Vulnerability Detection
- The model found more than 500 bugs by reasoning about the code like a human researcher, outperforming fuzzing techniques
- Focused on protecting open source software, and patches are now available; The company urges quick action while AI can still operate at scale.
Anthropic has released Claude Opus 4.6, the latest version of its most powerful large language model (LLM), and says it is “markedly better” at finding high-severity vulnerabilities compared to previous models. In fact, Opus 4.6 has so far managed to find more than 500 bugs of this type.
Anthropic said Opus 4.6 stood out for the way it found vulnerabilities “out of the box, without task-specific tools, custom scaffolding, or specialized prompts.”
He also added that unlike fuzzing, which is a standard vulnerability hunting technique, Opus works by reasoning about the code “as a human researcher would,” meaning it searched past fixes to find similar bugs that weren’t fixed, detected patterns that tend to cause problems, and understood the logic “well enough to know exactly what input would break it.”
Fixing open source software
Even using it on some of the best-tested codebases, projects that had fuzzers against them for years, Opus still managed to find high-severity flaws that went undetected for “decades.”
Anthropic said it began using Claude to help fix flaws in open source software, primarily because it runs “everywhere, from enterprise systems to critical infrastructure,” and because the vulnerabilities in this software are felt across the Internet. Additionally, much of the popular open source software used today is maintained by a small team of volunteers, meaning resources are more than limited.
The results have been notable, the company says: “So far, we’ve found and validated more than 500 high-severity vulnerabilities. We’ve started reporting them and are watching our initial patches arrive, and we continue to work with maintainers to patch the others.”
He concludes that AI models can now find high-severity vulnerabilities at scale, but emphasizes that this may not be the case in the near future. “Now is the time to act quickly to empower defenders and protect as much code as possible while the window exists.”
The best antivirus for all budgets
Follow TechRadar on Google News and add us as a preferred source to receive news, reviews and opinions from our experts in your feeds. Be sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form and receive regular updates from us on WhatsApp also.
