Signal EU AI Act defines prohibited uses and assigns responsibility to developers and deployers

Summary

This signal highlights the European Union's liability-focused approach to regulating AI. The EU AI Act tries to spell out which uses of AI are unacceptable, with social credit systems given as an example. It also sets out who carries responsibility when AI causes harm, generally placing it on the developers and deployers of systems rather than on end-users. The report contrasts this with jurisdictions such as California that emphasise transparency instead. The difference illustrates how responsibility for the same AI harm could rest with different parties depending on where it occurs. It also suggests that companies may need to adapt products to each regulatory regime.