Ethics Directive

The Standard

The register · live

Can anyone certify you as an AI ethicist yet?

Not quite. EN 18274 is due to be formally published by CEN in September 2026. Bodies are lining up ahead of it. An advertisement is not an accredited scheme, and telling the two apart is the entire job.
What we can and cannot confirm · last updated 1 Sep 2026
  • EN 18274 due to be published by CEN (26pp)expected September 2026, not yet confirmed
  • Standardisation mandate M/593confirmed
  • BSI, Bureau Veritas, DNV, Infocert listed as certifierssponsored listing
  • An accredited EN 18274 scheme you can sit todaynot confirmed

Those bodies appear under “get certified”, but as paid listings. A sponsored listing tells you a body is interested, not that a national accreditation body has approved a scheme. When one is genuinely live, this line will change, and it will be dated.

The standard

EN 18274: competence requirements for professional AI ethicists.

A European standard that sets out what a competent AI ethicist should know, be able to do, and hold to, expressed as knowledge, skills, and attitudes, and grounded in European values and fundamental rights. Here are the public facts, kept current.

Reference
EN 18274Due to be published by CEN · expected September 2026
Developed by
CEN & CENELECJoint Technical Committee 21 · Artificial Intelligence
Mandate
M/593EU standardisation request
Classification
ICS 03.100.02Governance and ethics; HR; IT application

The standard, still in draft ahead of formal publication, is already careful about what it is. It calls itself “intended to serve as a basis for the development of a certification scheme.” A basis, not the scheme itself. That is the gap this whole page tracks.

The three domains

Domain 01

Knowledge

What an ethicist must understand: European values and fundamental rights, the AI lifecycle, and the line between what is legal and what is ethical.

Domain 02

Skills

What they must be able to do: manage trade-offs, mediate dilemmas, assess societal impact, and communicate risk to very different audiences.

Domain 03

Attitudes

How they must carry it: independence, confidentiality, the willingness to raise an objection, and to use whistle-blowing channels when needed.

Knowledge and skills answer whether someone can do the job. Attitudes answers a harder question: what they do when doing the job costs them something. A system can clear every legal and policy check and still be the wrong thing to ship, because compliance describes a floor, not a conscience. And when something does go wrong, it is usually the ethicist left holding it, whether or not anyone was obliged to act on what they said. That is why the standard does not stop at competence. It asks whether you would still say something when saying it is expensive.

What the role covers

The standard frames the ethicist as an adviser who guides an organisation through ethical risk across the whole AI lifecycle: managing ethical trade-offs, analysing policy and mediating dilemmas, raising awareness and delivering training, developing ethical frameworks and guidelines, and conducting societal and ethical impact assessments. It also expects the role to reach beyond individual decisions, helping embed those same values into how the organisation actually operates day to day, its culture, not just its policy documents. Our gap report is built around these same areas.

Related standards it leans on