Un Decalogue to govern theartificial intelligence, suggested by Productive Activities Committee of the Chamber, because if on the one hand AI "like any new technology that has entered history heralds a whole series of both opportunities and potential risks", on the other hand it requires "balance", which "depends on human intelligence, that is, our ability to govern progress”.
And therefore, at the end of an investigation resulting from 17 sessions and 82 hearings and condensed into a final document of 24 pages, the Montecitorio commission indicates some paths to follow.
How to govern artificial intelligence: here are the 10 key measures
1) In the survey, which focuses on our country, it is generally underlined how the rapid and continuous evolution of new technologies, including artificial intelligence, poses the need for a adjustment also to legal level e regulate. In this sense, the regulatory sandbox (controlled environment where supervised intermediaries and operators in the FinTech sector can test, for a limited period of time, technologically innovative products and services) "represents the most suitable model to regulate those multiple and changing phenomena that affect also the artificial intelligence sector, with the aim of promoting innovation, allowing the development of this technology on the national territory and encouraging the attraction of companies and investments in Italy".
2) And from this perspective it is necessary to encourage and accelerate the business digitalization process (in particular medium-small ones) in light of the fact that generative AI can generate in Italy - with the same number of hours worked - up to 312 billion euros of annual added value, equal to 18 percent of the Italian GDP.
3) At the same time we hope for a training in terms of AI starting from primary school with a view to an idea of permanent training that keeps up with technology and therefore with the times.
4) The survey also highlighted the need to adopt a integrated approach to the development of banks personal data e systems of artificial intelligence - in particular through the development of open source models - so as to reduce dependence on technologies imported from large global players. In this sense it was noted that the data provided by public institutions must also be easily accessible, complete, and in machine-readable formats.
5) Again: constitute protected regulatory spaces for the entry into the market of new operators who can experiment with AI technologies without suffering immediate market competition.
6) The is then indicated constitution di a national authority ("decisive for the implementation of the new European legislation (the so-called AI Act)" with tasks of development, implementation, coordination, monitoring and control, also aimed at avoiding the danger that AI becomes a tool suitable for implementing or facilitate illicit activities.
7) The establishment of a has also been proposed institutional table, a sort of "control room" with the Government and the social partners for a general and continuous evaluation of the phenomenon, of its impact, especially in terms of employment and transformation of organizational models, which takes into account all the dimensions of the politics of digital transformation, therefore not only of technological and skills profiles, but also of ethical aspects and sustainable development objectives.
8) To streamline and speed up the implementation of technological innovation strategies that use AI, the proposal to establish a foundation or a company a capital entirely public that can facilitate the grounding of such policies.
9) Then there are the issues linked to the maintenance of jobs: to the foreseeable phenomena of unemployment frictional, in fact, are flanked by opportunity given by the search for new professionalism. The question is, therefore, “how we will act to accompany and protect work, where it will be allocated and whether poor or quality employment will be produced. Therefore, an adequate policy strategy is needed which sees the State acting as regulator, user and orchestrator of AI, promoting digital literacy at all levels of society, both in education and in businesses, promoting collaboration between the public and private sectors. Policies that can minimize job displacement while improving economic growth." Furthermore, today, for the first time in history, so-called intellectual work is also being questioned. Therefore, regulation aimed at protecting the copyright of works generated with the help of AI is also necessary.
10) Last, but not least, the question of computer security, a necessary prerequisite for the resilience, privacy, fairness, effectiveness and reliability of artificial intelligence systems. It is also fundamental - underlines the document of the Productive Activities commission of Montecitorio - to safeguard the privacy and rights of citizens (who are also consumers), especially with regard to the most vulnerable segments of the population (primarily the elderly and individuals with minimal digital skills).
