Article 1: Spotlight on Italy

This article marks the beginning of a series exploring the impact of AI on the film industry. Each piece will explore recent developments and their implications within a specific country's legal framework. In this inaugural article, we focus on Italy, examining how AI is transforming its cinematic landscape.

AI in the movie industry

AI already is – and will increasingly be – adopted in every stage of content creation, from pre-production to marketing and distribution. It was only a month ago that news broke of the first partnership between a major US entertainment company and an AI startup. The aim of the partnership is to develop a customized AI model based on the studio's extensive content catalogue. 

So, like for any emerging technology, AI's impact is twofold: part of the industry is already exploring AI's potentials (above all, cost saving), while another part is fighting against its possible downsides, like its impact on jobs and copyright infringement. But if you're wondering how AI is used in the audiovisual industry, here are some of the most relevant examples:

• Script analysis: AI can analyze vast amounts of scripts and predict their potential for success, so screenwriters can refine their work and producers can make more informed decisions. Of course, this isn't without drawbacks. AI will keep choosing scripts that are similar to something that had success in the past and its decisions may perpetuate biases.
• Casting and location scouting: AI can streamline the casting process by analyzing facial expressions, voices, and physical characteristics to match actors with suitable roles. Examples of this technology already exist: for instance, Casting Matching Process AI (CMP-AI) creates the character profile by analyzing it directly from the script, based on keywords using three narrative elements: physical-aesthetic, psycho-aptitude and scenic. Also, it can aid location scouting by analyzing hundreds of images or providing insights about weather predictions, crowd control and logistics. 
• Creating raw material: like in the partnership mentioned above (whose technical details have not been disclosed yet), AI can learn from large amounts of video and proprietary audiovisual material to generate, through latent diffusion techniques, new cinematic video sequences that will have to be refined. The advantages for the producer are obvious: reduced production time, cost containment and the possibility to test visual ideas in pre-production.
• Visual effects and post-production: AI is likely to become crucial in the post-production stage, both for small creators, who may be enabled to achieve visually stunning effects without the need for expensive equipment or big studios, and for professionals, who can significantly reduce costs and manpower. Although technical details about the model's architecture have not been disclosed, the idea is that, using AI, the studio can reduce production time, contain costs and test visual ideas in pre-production without mobilizing expensive resources.
• Marketing and distribution: by analyzing audience data, AI can tailor marketing campaigns, identifying which trailers or posters will resonate with viewers. It can also optimize distribution strategies by determining the ideal release dates and platforms for a movie. Some media companies have already signed deals with AI-based platforms able to provide insights and predictions on the success of movies and box office receipts. AI can also be used to increase fan engagement by creating a digital world around a movie, from character-like avatars to a variety of digital tools tailored to the movie's storyline.

A look at the Italian market: First pieces of legislation addressing AI

Last year’s protests in the US highlighted the potential dangers of AI for the film sector, such as the unauthorized use of authors', artists', and actors' works, voices, and images by generative AI. As a result, every country where the film industry plays a significant economic role is now grappling with how to balance the growth of the digital economy with protecting human creativity.

In Italy, the rapidly evolving influence of AI in the  industry can be already observed at the political and legislative level.

At the 81st Venice Film Festival, the Directorate for Cinema and Audiovisual industry of the Ministry of Culture organized, in collaboration with the European Commission, a round table with industry leaders titled "The Role of AI in the Audio-Visual Industry: Opportunities, Challenges, and Implications for Copyright and Artist Protection."

As expected, one of the primary concerns is how AI uses copyrighted material as input for training its models. The 2019 Directive on Copyright in the Digital Single Market introduced exceptions for text and data mining (TDM), allowing rights holders to control or exclude their content from these processes. The proposed AI Act builds on this by requiring AI model providers to respect copyright and provide detailed summaries of the data used for training. But a recent landmark judgment issued by a German Court (Kneschke v. LAION), which held that a nonprofit organization copying images didn't infringe copyright law, may have significant impact in the application of the TDM exceptions. For example, according to the ruling, creators cannot opt out of having their work used for AI training in scientific research.

Stakeholders are also discussing the protection to be assigned to the output of AI generated content. According to European copyright law, content generated entirely by AI, without human input, is generally excluded from copyright protection. But the debate continues over whether works created with AI assistance should receive some form of copyright protection. 

The draft law on AI approved in April in Italy proposes to revise national copyright law by clarifying that protected "intellectual works" must be of "human" origin and that AI-assisted works may still qualify for protection if their creation derives from the author's intellectual work. The draft law also aims to complement European legislation by facilitating the identification and recognition of AI systems in the creation of textual, photographic, audiovisual and radio content. 

In the new Tax Credit Decree, the Ministry has also introduced rules in relation to the eligibility of above-the-line costs related to using AI in relation to creative and artistic performances.

Industry reactions: Voice actors' new collective bargaining agreement

The film industry is already adapting to the challenges posed by AI. In December 2023, a new national collective bargaining agreement for voice actors was signed in Italy, introducing specific regulations concerning AI. In particular, the agreement sets forth the following rules, which apply until explicit agreements are made:

• Rights to reproduce a voice actor’s performance are limited to the specific audiovisual product for which the recording was made (eg the movie's trailer).

• The unauthorized use of voice data for AI development, including text and data mining, is expressly prohibited.

Further, legal disputes have also emerged between collecting societies representing voice actors and streaming platforms. These societies argue that platforms fail to fully disclose data and don't provide "fair and proportionate" compensation to voice actors. These ongoing disputes highlight the complexities of integrating AI into the creative industries while safeguarding artists' rights.

Conclusive remarks

The integration of AI into the film industry, both globally and in Italy, is both a promising and challenging development. 

AI offers tools for streamlining various aspects of film production. These innovations have the potential to reduce costs, enhance creativity, and tailor content to audience preferences. 

But the rise of AI also brings concerns regarding job displacement, copyright infringement, and the ethical use of AI-generated content. 

Italy, alongside the broader European market, is starting to actively address these challenges. While legislative measures are a crucial step, it will be equally important for the industry to raise awareness and implement best practices, such as those recently published by an international documentary producers group on how generative AI can and should be used in the genre.

Stay tuned for our upcoming spotlights as we continue to unveil how AI is shaping the film industry in different countries around the world.