Ethical AI Marketplace · public listing

Who in Portugal
knows where they stand —
and can prove it.

The organisations on this list did three things almost nobody has: they assessed their AI system against the text of the AI Act, established what risk tier they operate in, and authorised us to say so publicly. This is not a mark we sell — it is the result of a test anyone can repeat in five minutes.

The national picture · aggregate

The figure the country
was missing.

Every completed test is a data point. That includes those who did not authorise public mention: the aggregate never carries names.

Distribution by risk tier

    Assessments by sector

      What this chart is worth, and when

      A self-selected sample is not a representative one: anyone who takes a compliance test on their own initiative is, by definition, already above average in regulatory awareness. This aggregate measures those who assess themselves, not the universe of Portuguese companies.

      We say so up front, because the opposite — presenting this as a portrait of the country — would be exactly the kind of convenient reading an observatory cannot allow itself. The value is in the series: as the points accumulate, the distribution by sector and by size becomes comparable over time, and nobody in Portugal has that.

      How you get on this list

      You do not buy your way in.
      You take the test.

      • Complete the classification test 5 minutes
      • Authorise public mention at the end Optional
      • Demonstrate 3 of 5 governance pieces Criterion
      • Pay to be listed Does not exist

      The criterion is declared governance, not risk tier. A well-governed high-risk system counts for more on this list than a chatbot with no inventory — because that is what the AI Act measures, and what a customer needs to know.