While in Europe the approval process of the AI ACT has almost come to an end, in Italy the attention toward the potentials of Artificial Intelligence continues to grow.
At national level, on the one hand the Italian legislator has begun to worry about the risks connected to the use of AI systems in the editorial and information field, on the other, it acknowledged the benefits that AI could bring to institutional activities.
DDL 917/2023 “Measures on transparency of the content generated by artificial intelligence”
Last October 19th 2023, a law proposal was presented to the Italian Senate with the aim to impose a greater transparency on contents generated by AI systems. The text of the proposal, composed by only 3 articles, aims at safeguarding the right of users to a correct and truthful information, trying to regulate the diffusion of AI generated contents.
On this regard, the DDL requires “all AI-generated editorial contents” to be marked with labelling systems (watermark) as to make it clear to the users that they are seeing a partial or wholly AI generated content. This labelling obligation will burden on all subjects responsible of the disclosure or diffusion of such content.
The real methods of actuation of the labelling obligation are indicated in a specific regulation, which shall be adopted by the Italian Communications Regulatory Authority (in Italian Autorità per le garanzie nelle comunicazioni – AGCOM). The same Authority should then define the reporting and removal tools of content diffused in violation of the new law and the penalty regime to apply.
Even if this initiative is motivated by the understandable need to prevent abusive derives of the use of AI systems, it seems however to be addressing the issue from a rather reductive perspective. As a matter of fact, considering the complexity of AI systems, the variety of their practical applications, the global and interconnected dimension of the circulation of the content on the web, it seems difficult to be able to regulate the issue exclusively on national level. In this scenario, the job that the DDL would like to delegate to AGCOM – which should not only discipline the diffusion of all content, fully or partially generated by AI, but also repress and fine whoever publishes or discloses this type of content without watermark – seems definitely challenging.
Call for proposals for the use of generative AI for the Chamber of Deputies of the Italian Parliament
On February 14th 2024, the Italian Chamber of Deputies addressed the research and university world to open a dialogue on the possible uses of AI in the context of parliamentarian works. Conscious of the possible opportunities of these new technologies, the Camera expressed its availability to evaluate new tools and new proposals relating to 3 specific needs:
- Retrieve and organize qualified information, with the goal of making available to the Camera’s internal structures tools to support the preparation of parliamentary documentation, thereby optimizing internal work processes;
- Support the preparation of acts of legislative initiative, address and control, offering to deputes a system that can facilitate their activities drawing from the different types of sources, recreating a normative context, circumstances and facts of interests in the parliamentarian acts.
- Know the activity of the Camera, that is to make parliamentarian documentation and information about the activity of the parliamentarian organs more accessible, open and transparent for citizens, promoting a faster and more in-depth comprehension of parliamentarian works and legislative documents.
Proposals may be presented up until May 31st 2024 and the winning ones may be used as starting points to develop solutions to apply AI in the Italian Chamber of Deputies.
The authors of the winning proposals, who will be announced within July 2024, will also be awarded a cash prize, amounting to €10,000 for proposals related to scopes No. 1 and 2, and amounting to €5,000 for those related to the third scope.
Ilaria Feriti