The EU’s white paper on #AI claims that the introduction of “national initiatives” risks “weaken[ing] citizens’ trust”. Why so? Is it because citizens trust the EU more than their own governments? Or because the EU is more trustworthy than their governments? “Trust” is slippery..

The white paper goes on to say that building an “ecosystem of trust” requires ensuring compliance with “EU rules”. Why? Might not such an ecosystem require serious modifications to existing EU rules in light of observation of their effects?

“Trustworthy AI” is characterised as “AI based on European values and rules”. Can readily see how rules can be European, eg as part of EU law. But in what sense are values “European”? Genuine values are not jurisdictional, or is this a reference to a Europe-wide value consensus?

If the reference is to a mere consensus in value, there is good reason to question its normative force, not least at a time of rising authoritarianism in Europe.

Quite striking, but in line with talk of “European values”, that the white paper refers to striving to “export its values across the world”. Rather one-sides and inconsistent with the thought that ethical dialogue is a process with learners on both sides (Gadamer).

*one-sided