1 in 3 GitHub PRs now involve AI agents; new classifier doubles secret-leak prevention
GitHub Blog AI/ML · rss · 2026-10-08
GitHub shares nine quarters of data on secret leaks in the agent era: 1 in 3 PRs now involves an AI agent (up from fewer than 1 in 10 a year ago), and a new secret hits public code roughly every two seconds. Yet per-push leak prevalence shows no rising trend, and developer overrides of push blocks fell from 6.63% to 3.93% — developers are being outpaced, not careless. GitHub built a fine-tuned classifier with Microsoft Applied Sciences to extend push protection to unstructured secrets, assessing candidates in under 2ms and potentially more than doubling prevented leaks.
More from coding & agent
- A Developer's Map of the Data Tooling Landscape: Ingestion, Storage, Modeling, Governance — bibryam · 2026-10-08
- ChatGPT Work mode vs Codex: same quota, far more tasks done per 5-hour window — sasik520 · 2026-10-08
- Redditor complains Sonnet + Haiku agent swarm wrecks codebase whenever Opus budget runs out — MucilaginusCumberbun · 2026-10-08
- How to build an ad factory with Opus 5.5: auto-pause losing Meta ads and brief new ones daily — fekdaoui · 2026-10-08
- Kernaut: coding agents design Gaussian process kernels via program search — sirbayes · 2026-10-08
- Gartner: 70% of enterprises will abandon vendor-built agentic AI by 2028 — alex_verem · 2026-10-08