Quality Redesigned – Thinking in AI

What does the EU AI Act actually mean for a quality team already using AI tools day to day? That question sat at the centre of an impulse workshop I co-hosted with Stefan Schweizer at TQU GROUP — and the answers are just as useful whether or not you were in the room.

Why the EU AI Act matters for quality managers right now

The moment AI helps assess complaints, 8D reports, or FMEAs, one question comes up fast: what falls under the EU AI Act, and what does that mean day to day? The most useful takeaway: it rarely comes down to complicated legal text — usually it's two simple questions. Do your people know that and how AI contributes to the process? And does the technical decision stay traceably with a person? Answer those two cleanly in your own process, and you've already covered the essentials.

Practice over theory

What worked best in the workshop wasn't explaining AI — it was hands-on work with concrete tools, from structuring a complaint to assessing an inspection report. That's the same principle behind our work at KIundQ: AI only becomes genuinely useful once a quality team sees, using its own example, how a tool like an AI agent arrives at a traceable assessment step by step.

What you can take away directly

• Start where the pain is concrete — recurring complaints, for instance — instead of rolling AI out everywhere at once.
• Keep the technical decision with people. AI supplies the structure and the suggestion, your team supplies the judgement — which usually answers the EU AI Act's human-oversight question in practice.
• Ask early how an AI tool would be classified before rolling it out broadly — it saves rework later.

Thanks to TQU GROUP for hosting, and to every participant for the tough questions — I took away at least as much as they did. More impressions in the LinkedIn post on the workshop.

See how these principles show up in a concrete tool: FMEA Moderator.