Repository navigation
[uk ai resilience] UK AI governance: open-code risk report 2026-10-12 #68040
Closed
Replies: 1 comment
|
This discussion has been marked as outdated by UK AI Operational Resilience. A newer discussion is available at Discussion #68741. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Executive summary
Lookback 7d (since 2026-10-05): 354 commits (224 security-signal), 7 open security issues, 247 open code-scanning alerts, 0 secret-scanning alerts. Risk is dominated by supply-chain determinism in generated lock files and 3 critical Go quoting alerts. Sub-agents were not dispatched; analysis was done directly from the pre-computed data (limitation: lower confidence on ownership/runtime controls).
Asset graph summary
pkg/workflow/) – emits workflow YAML; 3 criticalgo/unsafe-quotingalerts (evals_steps.go:576, maintenance_workflow_yaml_jobs.go:586, :619)..github/workflows/*.lock.yml) – 227 alerts, mostly non-deterministicnpm install/uv pip install.Tier classification
Control verification gaps
make recompile(good patchability); release provenance gap per Add per-binary attestations to release artifacts #52302.Risk scoring
Remediation queue
Exception register
None.
Operational metrics baseline
All reactions