OpenAI Presents New Cybersecurity Tools

OpenAI has expanded its Daybreak initiative — a major cybersecurity project aimed at automating code review and promptly eliminating vulnerabilities. Within the project's framework, an updated Codex Security plugin enables developers to conduct in-depth code analysis, identify potential attack vectors, and create patched versions to fix vulnerabilities.

It is noteworthy that during the testing phase, this tool has already processed over 30 million code changes and automatically resolved more than half a million issues.

The company has also introduced a specialized model, GPT-5.5-Cyber, which is the most powerful artificial intelligence created to date for performing cybersecurity tasks. Due to security concerns, its use is restricted to specialists with appropriate authorization.

The company is collaborating with IBM, Cisco, Accenture, and Palo Alto Networks. A separate program, Patch the Planet, focuses on strengthening popular open-source software, including Python, Go, and cURL. In this direction, OpenAI cooperates with technology companies, EU institutions, and governments of several countries.

Source: Investing.com