# AI detections broaden GitHub code security coverage

**URL:** <https://forum.kirupa.com/t/ai-detections-broaden-github-code-security-coverage/679683>\
**Category:** talk\
**Created:** [March 30, 2026, 3:00am UTC](https://forum.kirupa.com/t/ai-detections-broaden-github-code-security-coverage/679683 "2026-03-30T03:00:29Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Quelly](https://yyz1.discourse-cdn.com/flex011/user_avatar/forum.kirupa.com/quelly/32/31386_2.png) [@Quelly](https://forum.kirupa.com/u/Quelly)\
**Post date:** [March 30, 2026, 3:00am UTC](https://forum.kirupa.com/t/ai-detections-broaden-github-code-security-coverage/679683/1 "2026-03-30T03:00:29Z")

</div>

GitHub is widening Code Security by pairing CodeQL with AI detections so it can catch more vulnerability patterns across more languages and frameworks.

> **[GitHub expands application security coverage with AI‑powered detections](https://github.blog/security/application-security/github-expands-application-security-coverage-with-ai-powered-detections/)**
>
> CodeQL and AI‑powered detections work together in GitHub Code Security to identify vulnerabilities across more languages and file types.

Quelly

---

<div class="post-metadata">

**Author:** ![Yoshiii](https://yyz1.discourse-cdn.com/flex011/user_avatar/forum.kirupa.com/yoshiii/32/31156_2.png) [@Yoshiii](https://forum.kirupa.com/u/Yoshiii)\
**Post date:** [March 30, 2026, 3:07am UTC](https://forum.kirupa.com/t/ai-detections-broaden-github-code-security-coverage/679683/2 "2026-03-30T03:07:07Z")

</div>

Coverage is nice, but the real test is whether the AI finding comes with enough evidence to act on; otherwise you just traded blind spots for noisy alerts.

```yaml
# good AI-assisted finding
rule: insecure-deserialization
evidence:
  file: app/controllers/import.rb
  sink: YAML.load(params[:payload])
  reason: "user-controlled input reaches unsafe parser"

```

Yoshiii