Hardly a film industry event passes these days without a panel discussion addressing the opportunities offered to the creative community by the deployment of artificial intelligence (AI) and the impact of this digital revolution on the future development cinema.
This spring saw the staging in Cannes of the World AI Film Festival (WAIFF) which highlighted a new era in cinema where artistic vision meets AI-powered creation, and is now to be followed by a series of international editions around the globe including Berlin, Los Angeles, Seoul and Sao Paulo ahead of WAIFF’s 2027 edition next April.
In conversation with German Films Quarterly, Dr. Gerd Hansen, a partner in the Munich office of the international law firm Fieldfisher, gives, among other things, an insight into the current legal issues regarding the deployment of AI in Germany’s film industry and suggests measures that the nation’s film funding bodies might consider for projects wishing to use generative AI.
Dr. Gerd Hansen © Fieldfisher
Dr. Gerd Hansen
An experienced entertainment lawyer with a deep passion for media content, Dr. Gerd Hansen, a partner in the Munich office of the international law firm Fieldfisher, specialises in production legal services covering all phases of the content production process and also offers legal advice concerning the use of generative AI within the entertainment sector.
How would you assess the current state of play as far as the use of artificial intelligence in the German film industry is concerned? Is the industry sufficiently prepared, both legally and structurally?
Dr. Gerd Hansen: The film industry has always been renowned for its ability to quickly embrace technological innovations. We are seeing the same thing happening now. My impression is that generative AI has already penetrated deeply into all phases of the production process, from development through production and post-production to distribution. Although Germany may have missed the boat when it comes to AI models in an international comparison, I still see a great opportunity for Germany to be at the forefront when it comes to the implementation of AI.
Which legal issues relating to AI are currently occupying producers and filmmakers the most?
The focus In my consulting practice is currently on AI risk management – in other words, ensuring that AI is used correctly. The questions are very diverse and touch on various different areas of law such as copyright, data protection and, increasingly, regulatory frameworks such as the European AI Regulation or the AI collective agreement. A very hot topic at the moment is the obligation to label deepfakes, which comes into force on 2 August 2026. A great deal is still in flux here, and I am really enjoying being able to help shape this emerging area of law.
Where, in your opinion, are the greatest legal uncertainties at this point in time?
Clients are increasingly coming to me because they expect a sort of ‘AI Production Readiness Check’: to avoid AI-related risks, secure eligibility for funding, maintain insurability and ensure commercial exploitability. At the moment, I sense a great deal of uncertainty in the market, particularly regarding the AI labelling obligation. It’s important here to establish a pragmatic approach to the European guidelines. For we could then run the risk of bringing about ‘banner blindness’ if even the slightest AI-assisted alteration leads to classification as a deepfake that needs labelling. AI transparency, yes – particularly in the field of journalism – but please with a sense of proportion when it comes to fictional productions.
Is Germany by international comparison a pioneer, somewhere in the middle, or lagging behind when it comes to the legal regulation of AI in film?
I would definitely regard Germany as a pioneer in this area. With the AI collective agreement negotiated in 2024 – which I had the privilege of helping to finalise as a legal adviser on behalf of the German Producers Alliance – we established guidelines for the use of AI in acting at a very early stage by international standards. These provide first movers with the necessary legal certainty and, at the same time, give actors the assurance that they will not lose the rights to their likeness if a digital replica is used in a production. In some respects, the regulations are even more progressive than the guild agreements with the US actors’ union SAG-AFTRA.
How should German film funding bodies handle projects that use generative AI in screenwriting, image composition, animation, voice-overs or post-production?
I get the impression that funding bodies are, in some cases, gradually and very cautiously opening up to this. For example, Bavaria’s Minister for Media, Florian Herrmann, recently announced at the Munich International Film Festival that, as part of the ongoing update to the FFF Bayern’s guidelines, an AI funding component would also be included – which I feel is a very positive thing. It’s constantly claimed that AI will now make everything cheaper and quicker to produce.
However, we are still in an early phase of innovation with high development costs. The use of AI in professional moving-image production still involves additional experimental effort. AI costs time and money. So, I’m advocating, amongst other things, for project-based funding for AI prototypes – in other words, film funding for AI that supports first movers with seed funding to evaluate the feasibility of using AI.
Are special funding criteria needed to ensure that human creatives can continue to play a significant role in funded films?
To begin with, I’m convinced that filmmakers of flesh and blood will continue to play a very central role. At the same time, particularly given the international competition between film hubs, I believe it would be wrong for the film funds to aim at artificially restricting the use of AI. Blanket bans on AI won’t be helpful. They would ignore the reality of the production processes. The challenge is rather about ensuring a nuanced, transparency-based approach to the new technical possibilities. This is where disclosure requirements and qualitative funding criteria can come into play. The film funds could, for example, require applicants to have an appropriate AI risk management strategy in place to ensure responsible and legally compliant use of AI. This may include producers demonstrating that they have taken all reasonable steps to ensure that the crew possesses a sufficient level of AI competence according to Article 4 of the AI Act. Such an AI risk management strategy could also provide for appropriate organisational measures to raise awareness of AI-related risks, as well as processes to mitigate legal risks. These include, in particular, measures to safeguard the copyright, neighbouring rights and personal rights of third parties, so as to minimise the risk of AI output infringing the law and to ensure that the work of human creators is respected. It would also make sense to require compliance with relevant industry agreements, such as the AI collective agreement. What’s more, I think that accompanying measures are in fact even more important: AI upskilling programmes, not least also for established filmmakers. The goal shouldn’t be to protect people ‘from AI’, but rather to empower them to use the technology skilfully.
Can a film that’s been produced predominantly using generative AI still be regarded as a cultural work and as a German film eligible to receive funding?
I believe that even a feature film produced predominantly with the aid of AI can, in principle, pass the cultural test, provided that the corresponding criteria are met and the creative process is significantly shaped by human input. The cultural quality of a film does not depend on whether certain technical tools are used, but rather on which story is being told, the connections the work has to German culture and German subject matter, and whether it is being produced with actors and other film-makers from Germany. A film doesn’t lose its cultural character simply because generative AI was used in its production.
The cultural test already takes criteria into account such as subject matter, language, location, cultural references or the involvement of creatives – and not whether a particular cam–era, editing software or, in future, an AI tool was deployed. In my opinion, it would therefore be misguided to make funding eligibility contingent across the board on the use of a specific production technology.
What might a legally sound, yet unbureaucratic AI guideline for German film funding bodies look like?
Answering this question would certainly go beyond the scope available here. A workable AI funding guideline should, in any case, not repeat the mistakes of some overly cautious AI guidelines which threaten to stifle innovation through strict AI tool requirements or excessive transparency obligations. I would also like to see a certain openness towards new, innovative formats such as vertical micro-drama series where the use of AI is virtually inevitable. In my view, imposing a blanket requirement for a broadcaster’s or distributor’s contract, for example, would fail to do justice to the scale of the upheaval that’s currently taking place.
How meaningful and important are specialised AI film festivals such as the World AI Film Festival in Cannes?
I consider specialised AI film festivals to be highly relevant during the current transitional phase. Such events serve to build community, facilitate the transfer of expertise and provide reassurance to an AI creator scene that is currently setting out to revolutionise film production around the world. What I find fascinating is how the “traditional” Cannes Film Festival has (so far) distanced itself from this. This was also almost physically palpable this year in the “Innovation Village” which was physically separated from the rest of the market and effectively pushed to the edge of the harbour basin. I agree with Demi Moore, who said in Cannes: “AI is here. To fight it is a battle that we will lose.”
Where do you see the most sensible and legally least problematic applications of AI for the film industry?
The areas of application that are legally the least problematic are those in which AI is used merely as a tool under human control and where the AI output doesn’t then appear in the visuals or audio of the final production. The development phase, for example, can include AI-assisted research, pre-visualisation using mood boards, storyboard drafts, automated shooting schedule optimisation, budget calculation and casting support. And then, of course, at the stage of post-production: colour grading, upscaling, subtitling, automated transcription, etc.
What are now the most important legal steps filmmakers and producers should be aware of before using an AI tool?
Anyone who rushes in headlong is bound to come unstuck. So I recommend that producers adopt a structured approach to AI risk management. Before using new AI tools, they should always follow a so-called ‘whitelisting’ process using a standardised checklist in order to avoid any unpleasant surprises.
Angela Sonntag and Martin Blaney
