Tag: #vuln-detection7 results
Microsoft is integrating Anthropic's Mythos into its security development program. This AI tool will help identify vulnerabilities more quickly. It also enables faster development of fixes for software security issues.
OpenAI announced limited release of GPT-5.4-Cyber, mirroring Anthropic's approach. The model detects security vulnerabilities in software. Access restricted to trusted companies.
OpenAI released the Cyber model to select users, specializing in software security vulnerability detection. It competes directly with Anthropic's Mythos tool, launched a week earlier.
Anthropic debuts Claude Mythos Preview in Project Glasswing cybersecurity partnership with Nvidia, Google, AWS, Apple, Microsoft. The model flags vulnerabilities autonomously in every major OS and browser. Not planned for public release due to security concerns.
Anthropic’s Claude Opus 4.6 identified 22 security flaws in Firefox during a two-week test with Mozilla, including 14 serious ones. It analyzed the JavaScript engine first, then nearly 6,000 C files, generating 112 error reports. Most vulnerabilities were fixed in Firefox 148, though Claude excelled more at detection than exploitation, succeeding with only two bugs after $4,000 in API costs.
Mozilla confirmed 271 vulnerabilities detected by AI tool Mythos with almost no false positives. Firefox developers have fully embraced AI-assisted bug discovery.
Codex Security does not use traditional SAST reports. It employs AI-driven constraint reasoning and validation to identify real vulnerabilities. This method significantly reduces false positives compared to conventional approaches.