type: incident · created: 2026-08-10 · updated: 2026-08-10 · tags: [incident, ai, patching, vulnerability, research] · confidence: high · affected_sectors: [technology, software-development] · au_impact: true
More Than Half of AI-Generated Security Patches Are Broken
New research reports that more than half of AI-generated security patches are broken — failing to remediate the vulnerability, introducing regressions, or breaking functionality.
| Attribute | Detail |
|---|---|
| Finding | >50% of AI-generated security patches are broken |
| Failure modes | Fail to remediate; introduce regressions; break functionality |
| Source | CyberScoop — Tier 2/4 |
A cautionary signal for teams adopting AI-assisted vulnerability remediation, underscoring that machine-suggested fixes still require human verification before production or OT deployment — directly relevant to the AI-assisted remediation strategies many Australian and NZ programmes are trialling.