Possible Risks for the Italian Artificial Intelligence Strategy in comparison with the EU AI Act

 

The following is a commentary, written by ESR11-Luca Nannini in Italian a few days after its publication at the end of November 2021, on the National Strategic Plan for Artificial Intelligence 2022-2024 drafted by the Italian Government through the Ministry of Economic Development, the Minister for Technological Innovation and Digital Transition, and the Ministry of University and Research.

As a preamble, are discussed the challenges inherent in holding a deontological approach and mature dialectic on AI ethics. The preamble briefly points out how, so far in the industry, AI governance issues lacked of prescriptive and independent audit systems as well as a mature debate. It also highlights the current limitation of the author’s personal knowledge in the legislative domain. This is due to the dynamism involved in the multi-disciplinary area of AI ethics where contributions from computer science, sociology, politics, and economics should be ponderated.

In the commentary are advanced potential risks that could arise in light of the socio-technical implications of the new EU AI Act in combination with other international geopolitical dynamics. Three degrees of risk are individuated, namely the international AI market, compliance within EU AI legislations, and Italian AI readiness.

The international AI market is signaled by US & China’s interest in developing AI projects within EU stakeholders and their implications. In particular, the dependence on external AI frameworks and the threat of draining the Italian and EU digital workforce might jeopardize the Italian strategic investments in AI R&D. Within EU AI legislations, the Italian strategy is well-aligned. But given the uncertainties of the current EU AI Act, Italy might suffer more the sociotechnical AI implications compared to the most developed EU countries, since the current national AI R&D and industry market cap lacks behind. At the core of these ‘external challenges’ lays how well the Italian Government will foster practices of AI literacy and readiness to the single citizens in the short term.

To read the full bolg post please click here (available in Italian): https://www.privacy-network.it/possibili-rischi-per-la-strategia-italiana-sullintelligenza-artificiale-a-confronto-con-leu-ai-act/