AI-Generated Patches Could Ease Developer, Operations Workload

The report outlines how large language models are being applied to automate bug fixing and patch validation, with Google’s Gemini reportedly fixing a subset of sanitizer-detected issues and leveraging automated builds to filter bad patches. It compares general-purpose and curated approaches (e.g., Veracode) and explores extending AI-driven patching into IT operations to reduce patch risk and vulnerability backlogs, while noting ongoing challenges around accuracy, validation, and operational impact.