The EU AI Act’s High-Risk Rules Are Here — and They Demand Qualified Human Oversight

The European Union’s Artificial Intelligence Act has moved from statute to reality. The obligations for high-risk AI systems listed in Annex III — which include AI used in employment, worker management and access to essential services — became a central compliance focus for 2026, with core provisions scheduled to apply from 2 August 2026.

For organisations that deploy AI in safety-critical and workplace settings, one requirement stands out: effective human oversight. High-risk systems must be designed and operated so that real people can understand, monitor and, where necessary, override the system’s decisions.

Oversight is only as good as the people doing it

The Act is explicit that the individuals responsible for oversight must be competent — properly trained, qualified, and given the authority and capacity to intervene. A “human in the loop” is not a checkbox; it is a role that requires demonstrable skill in understanding AI behaviour, limitations and failure modes.

That is a significant shift. Regulators are no longer satisfied with organisations simply owning an AI system; they expect evidence that the people governing it know what they are doing.

A moving deadline, a fixed direction

The timeline itself has been debated. In late 2025 the European Commission proposed a “Digital Omnibus” package that would defer some high-risk obligations — including those covering employment — to December 2027, with rules for AI embedded in industrial machinery following in 2028. The dates may shift, but the direction of travel is unmistakable: qualified, accountable human oversight of AI is becoming a legal expectation, not a best practice.

With penalties for non-compliance reaching up to €35 million or 7% of global turnover, the cost of getting oversight wrong is substantial.

Why this matters for AI-safety competence

This is precisely the capability the AI Safety Council’s certifications are built to validate. Credentials such as the Certified AI Safety Professional (AISP®) and Certified AI Risk Management Professional (AIRP®) assess whether a professional can design controls, exercise meaningful oversight and evidence compliance against recognised standards — exactly the competences regulators now expect.

As the regulatory floor rises, verifiable professional competence becomes a competitive advantage: for individuals seeking to prove their readiness, and for employers who need to show that their people can be trusted to keep AI safe.