T he European Parliament has approved the AI Act, the first comprehensive law on artificial intelligence. First compliance deadline: 6 months. In two years all the rules will become operational. «The punctuality of the approval must certainly be underlined», highlights Franco Siddi, president of Osservatorio TuttiMedia . «Regulation can avoid the risks of manipulation in the field of information. Innovation must be supported, but it is undeniable that the encounter between man and machine must find a ground for development that allows the protection of human creations, which in any case have the dimension of uniqueness and therefore require certification".

For Siddi, human intervention is a guarantee of authenticity, transparency, it ensures the rights of those who create content, images, videos, and of those who use them. «I am thinking of the readers, the citizens, the spectators who», specifies Siddi, «must be able to know who produced the content. And then there are the rights of authors, writers, publishers, and the cultural industry as a whole. The approval of the European AI Act opens the way to the maximum possible transparency, in a time in which algorithms and innovation, in general, fascinate, but this does not mean we need to give in to the domination of the few" .

Siddi also points out that "technology evolves rapidly and training is already done with big data and web scraping techniques adopted to date". Finally, Siddi says that « we are only in a phase of evolution and the European AI Act is an attempt which is probably not sufficient on its own, because the big players are outside the old continent. The theme of fakes and the remuneration of companies for creative work must be central in the subsequent steps. It will therefore be necessary to develop global awareness policies. Meanwhile, Italy is trying to find its way, so much so that it has placed AI management as a priority among the central themes of the G7, which it presides over."

(Unioneonline)

© Riproduzione riservata