Back to @tgroenwals's timeline
tgroenwals shared this post · 1d ago
LinkedIn
Alex Miguel Meyer
Alex Miguel Meyer

Adding a human next to the AI is not oversight.
Regulators do not care that your AI is fast.
They care who signed off.

EU AI Act Article 14. NIST AI Risk Framework.
Colorado AI Act. Texas TRAIGA.

All four require human oversight with three elements.
Context. Authority. Rationale.

Most governance frameworks skip all three.
They add a checkbox. Not a checkpoint.

Here is how to deploy Human-in-the-Loop across the 5 governance layers.

→ Ethics:
Named reviewer with real veto authority.

→ Risk Management: Tiered checkpoints.
High-risk pauses. Low-risk runs.

→ Accountability:
One human signs off. One name in the log.

→ Transparency:
Cannot explain it? Do not deploy it.

→ Compliance:
Signed human approval before every deployment and update

The real question is not "do you have a human in the loop?"
It is "does that human have the authority to say no?"

If the answer is no, you do not have oversight.
You have theater.

What is one governance decision at your company that runs without a real human check?
⬇️ Let me know in the comments

→ Join AI-Empowered Leaders:
My weekly newsletter with actionable AI insights from my work as AI advisor, trainer & coach.
Sign up here 👇
<https://lnkd.in/eUmy2Bdp>

♻️ Repost to help your network build governance that actually works

Muneeb Ahmed One thing I've noticed is that meaningful human oversight isn't about inserting a person into the workflow. It's about making sure someone has both the information and the authority to challenge the AI's recommendation. Without those two things, the approval becomes procedural rather than a genuine control.
Muneeb Ahmed Alex Miguel Meyer Exactly. Both are necessary, but there's a third piece as well: accountability. When a decision can be traced back to a specific person with the context and authority to intervene, oversight becomes much more than a compliance exercise.
Tanzeela S. This aligns closely with ISO/IEC 42001. Effective AI governance requires clearly defined roles, documented decision-making, and accountability throughout the AI lifecycle. Human oversight should be a measurable control, not just a policy statement.
Alex Miguel Meyer Author Tanzeela S. Well said. Standards like ISO/IEC 42001 move human oversight from intention to execution by making accountability something you can actually demonstrate.