OpenAI introduced Aardvark, an AI-driven model aimed at automating bug hunting and patching. Utilizing advanced large language model capabilities, Aardvark scans source code repositories to find vulnerabilities and assess their severity, enabling timely remediation. The model can develop threat models and validate exploits, making significant contributions to security processes. Aardvark has already identified 10 Common Vulnerabilities and Exposures (CVEs) and can also spot logic and privacy bugs. Offering a free version to the open-source community emphasizes OpenAI’s commitment to enhancing security across ecosystems. This move aligns with the company’s updated vulnerability disclosure process, aiming to relieve pressure on developers regarding strict timelines.
👉 Pročitaj original: CyberScoop
