Dear AI-ACT co-rapporteur,
Dear Mr Tudorache, Member of the European Parliament,

We, the undersigned, Guy T'hooft and Joël De Decker, would like to thank you for the very clear answers you gave at the conference "Bridging the Gap: Artificial Intelligence and Fundamental Human Values" held on 15 November 2023 in the Library of the European Parliament.

As the time available to participants is very short, we are writing to provide you with the full text of our speech. We would be grateful for any comments you may have.

Statement by the European Association for Freethought (AEPL)

At the end of 2019, AEPL wrote a letter to the President of the European Commission, saying in part:
"Your European vision of artificial intelligence, giving primacy to ethics, has caught our full attention..... We fully support the objectives described in the "Ethical Guidelines for Trustworthy Artificial Intelligence" published in spring 2020, following the work of the high-level expert group set up by the European Commission."

However, we are concerned about a recent publication by an interest group stating the following thesis: "Europe will be left behind if it focuses on ethics and does not follow the development of AI". AEPL considers that this argument does not serve the interests of European citizens who are concerned about their own ethical values, in particular human rights.

The European vision of "ethics by design" confirms an ex ante approach, enabling ethical objectives to be set before research and technological development, rather than after the event.

Within the legislative framework of Artificial Intelligence, Europe has a duty to continue what it has achieved in terms of personal data protection: to be an exemplary conscience for the rest of the world.

We note that Parliament is particularly attentive to the aspect of facial recognition, which it wishes to ban completely. But here, as in other applications of Artificial Intelligence, it is not the algorithm itself that is dangerous, but the way in which it is used.

Chinese-style facial recognition, with its social branding, should be totally banned in the EU because it runs counter to its values. But facial recognition on our cameras, which triggers the shutter when the picture is taken, has nothing to do with the values of the European Union.

Finally, we would like to express our strong support for the following three points:

1) Artificial intelligence is characterised by the fact that it can, by construction, always be wrong, hence the ethical problems posed by the use of AI. It is therefore important that the various versions of AI-ACT are continually modified to take account of the rapid technological changes in the field of AI and the evolution of public opinion.

2) The worst possible outcome would be no inter-institutional agreement, which would mean the abortion of AI-ACT. Europe's citizens expect a social pact that addresses their fears, whether well-founded or imagined, about artificial intelligence.

3) As you rightly confirmed, educating the "man in the loop" is of paramount importance.

We remain at your disposal for any further information you may require.

Yours sincerely

 

Joël DE DECKER Guy T'HOOFT

Director of AEPL Chairman of AEPL

Leader of the AEPL AI group