The EU AI Act is one of the most significant pieces of AI legislation in the world. This is accepted, even if reluctantly, by all in AI. What is not often realised is that this is written as product safety legislation, a point of view that is unfamiliar to many in AI.

Like other product safety legislation, it is written in fairly general terms, using words such as “unbiased”, leaving it to standards to describe what this actually means in detail. Again, this is unfamiliar to many in AI. We will therefore first look at the Act, and the European standardisation process. In particular, we will look at the various actors in this process.

Then we will ask what this has to do with Symbolic Computation: it is the general view that “Artificial Intelligence” as commonly understood, i.e. Machine Learning, has nothing to do with symbolic computation. “General”, but not “universal”. In fact, there is a community applying some of the tools of symbolic computation to prove (as opposed to claim, occasionally supported by statistics) facts about certain AI applications. In particular, this community asks, and partially answers, the “who shall verify the verifications” conundrum that the standard symbolic computation paradigm is also facing.