Three proposals to strengthen EU AI Act

Trust thought for today - the importance of citizen inclusion and engagement in the scope and design of governance of Artificial Intelligence.

Ada Lovelace Institute gets more and more impressive. Here is a blog on their thoughtful proposals to strengthen EU law on AI. Congrats Alexandru Circiumaru, super job.

Three proposals to strengthen the AI Act

Agree with all of their recommendations, but particularly like the points about 'Ensuring that those who use, and are impacted by, AI are empowered to participate in its regulation', one of the things I bang on about in our Trust and AI governance but much more thoughtfully articulated and in-depth here.

He explains that the EU AI act talks of ‘users’ as the deployers of AI, not those who are affected by AI systems - ie all of us. The EU proposal does not create a ‘meaningful’ space for the participation of those affected by AI or consult them in any way in the shaping of the governance, standards, regulations, ethics. Neither does it give citizens any right to challenge or complain about AI systems in use when they go wrong and negative affects occur. ‘Algorithmic reparation’ I saw that called the other day.

If, as Margarethe Vestager said ‘trust is a must, not a nice to have’ for AI my research on trust and tech governance shows this inclusion, openness and accountability to society is totally essential.

If interested here is the blog Sam Brown and I wrote on that - the most read of my Trust thoughts for the day in 2021 which cover my actual work!

If trust is a must for AI governance what do regulators have to do:

https://lnkd.in/dkxYPVj2

One for you if not seen Alistair Nolan, Sebastien A. Krier, Vanja Skoric, Ekkehard Ernst, Anja Kaspersen, Wendell Wallach, Markus Krebsz, Kay Firth-Butterfield