Keynote: Using LLMs and generative AI to fix software vulnerabilities - Asankhaya Sharma

Description
Large enterprises struggle with software vulnerabilities: complex systems, slow patching, and poor risk prioritisation lead to security weaknesses, breaches, and data exposure. In this talk, we will show how one can leverage recent advances in LLMs and generative AI to create and validate private, secure and multiple file fixes that rival even expert developers. State of the art (SOTA) AI open-source models (like StarCoder) for code generation and assistance can enable enterprises to securely and privately remediate vulnerabilities in …

breaches data data exposure enterprises exposure fix generative generative ai keynote large llms patching poor private risk security slow software software vulnerabilities systems vulnerabilities weaknesses