analiza aiqbench

This article highlights a critical gap in AI-assisted cybersecurity where models like OpenAI's and Anthropic's Claude are used to identify and exploit software vulnerabilities faster than defenders can patch them. The analysis underscores that while AI accelerates the discovery of flaws in widely used systems like PHP and Linux, media coverage remains disproportionately focused on breaches rather than the unpatched vulnerabilities themselves, leaving infrastructure exposed for extended periods.