Written by Deniz Ölmez
In conversations about AI governance, the terms often sound big and abstract: fairness, transparency, accountability, robustness. But inside real projects, the work is far more grounded. It looks like testers trying to understand a model’s limitations. It looks like product owners asking what must be logged. It looks like engineers wondering whether an unexpected output is a bug or a governance breach.
Across the topics announced for this year’s Swiss Testing Day sessions, one observation becomes clear: governance isn’t something that happens far away in a legal office. It lives inside the daily routines of development teams – in version control, in test reports, in data selection, in decisions about what should never be automated.
As organisations adopt the EU AI Act and emerging standards like ISO/IEC 42001, testers find themselves in a new role. They become translators, connecting regulatory principles with practical quality criteria. Governance becomes visible every time a team documents why a model was retrained, or when someone asks whether a specific decision should still require a human review step.
In high-maturity teams, governance shows up not as a layer on top of the SDLC but as part of its rhythm. Risk classification becomes a natural part of requirements discussions. Model documentation evolves alongside the code. Testing includes not only performance but behaviour under drift, stress, and edge situations. And oversight becomes something that can be demonstrated, not just claimed.
What surprises many teams is that good governance often improves velocity rather than slowing it down. Clear responsibilities reduce debate. Better documentation speeds up debugging. Thoughtful oversight prevents incidents that would have cost weeks to recover from.
At its heart, governance is about building systems we can stand behind – systems whose decisions are explainable, traceable, and responsibly deployed. And QA sits at the centre of that effort, turning broad principles into everyday practice.
As AI becomes more deeply embedded in critical processes, the organisations that succeed will be the ones that treat governance not as compliance work, but as quality work – and testers will be among the first to lead the way.
Join Us at Swiss Testing Day 2026
Curious about how to bring AI into your testing processes? Join us on March 26th in Zurich for the Swiss Testing Day 2026! This is your chance to connect with industry leaders, share insights, and explore innovative solutions shaping the future of software testing.
Whether you’re an experienced tester or just starting out, there’s something for everyone. Plus, you’ll have the opportunity to network with professionals who are just as passionate about quality assurance as you are.
Limited tickets remaining.



