Strategic Advisor|Advisory Board Member|Author|Conference Speaker|Founder|Pastor|AI Security Researcher|Cloud Security Evangelist
This is indeed very interesting and can be extended to other area of security operations where excess toil can lead to the accumulation of “security debt”.
AI-Powered Patching? Really interesting and promising read from Jan Nowakowski and Jan Keller at Google The paper looks at their experiences leveraging AI to scale patching They utilized automations in a pipeline to prompt LLM's to generate code fixes for human review It utilizes their Gemini model to successfully fix 15% of bugs discovered in testing, leading to hundreds of bugs being patched. As they point out, this automated capability has big implications for more secure code and the LLM's will continue to improve across languages as time goes on. Love seeing the use of AI to make software more secure, among all the FUD about the threats of AI. #cybersecurity #ai #software
