At this year’s Marché du Film (14-21 May), Cannes Next and the International Screen Institute hosted a panel titled “Reimagining Production: Navigating AI Innovations, Emerging Business Ventures and Legal Landscapes”. The event brought together experts from across Europe to discuss how AI is reshaping the creative process, the industry’s business architecture and the urgent legal conversations that must follow. Moderated by Sten-Kristian Saluveer, the session struck a careful balance between pragmatic optimism and critical awareness.

The conversation opened with a lightly provocative question: “What keeps you positively awake at night?” For Verena Puhm, an Austrian-US AI creative director with roots in traditional filmmaking, the answer was simple: curiosity and speed. “When I entered this space, I felt like a child again. AI allows me to make more, faster, better. I’m 36 – I don’t want to make just five or six films in my life.” To her, these tools are not about replacing artists, but empowering them.

This notion of empowerment was echoed by Polish producer Marta Krzeptowska, who brought a refreshingly hands-on perspective. “I remember when my colleague said, ‘Why aren’t you using AI or ChatGPT for this?’” she recalled. “I had this difficult script with a very rough sex scene. I asked ChatGPT to make it ‘softer’ – and it nailed it instantly.” She now handles up to 12 projects at once, compared to just three, pre-AI. “It’s not less artistic; it’s accelerated. I can do so much more.”

However, Krzeptowska also sounded the alarm on authorship and creative protection. “We’re standing at the edge of a new century,” she noted. “It’s hard to legislate what’s coming. Even the darkest scenarios are possible now.”

Legal expert Charlotte Lund Thomsen, founder of IP Legal & Policy Advice, reinforced the necessity for clear legal frameworks. She works closely with the EU on the evolving effects of the AI Act, advocating for the audiovisual sector’s inclusion. “We need to ensure that policymakers understand we’re not just working with technology; we’re building culture, jobs and economic growth.”

Lund Thomsen stressed the importance of documenting the human contribution to the creative process, pointing out that AI-generated work cannot be copyrighted unless clear evidence of human input is provided. She cited Finland as an example, where artists are encouraged to keep a detailed record of each creative step. She also warned against feeding entire scripts into AI tools, noting that doing so effectively means training the model with one’s own intellectual property.

The discussion turned practical when Saluveer asked about workflow evolution. Puhm described a new kind of precision. “I write down each shot and identify the tools that best serve it. For breakdancing, for example, only one tool – Google Veo 2 – currently gets it right. It can take as few as 20-25 prompts to get a shot right, or even more than 200,” she revealed, admitting that the current level of technology still doesn’t allow for perfect results.

Importantly, she doesn’t see AI removing jobs, but rather reshaping them. “Many VFX artists have become the best AI artists. They already know how to problem-solve through colour correction or compositing. It’s the same mindset.” She also offered a cautionary note: “Don’t prompt in the style of a famous director or reference existing IP. That’s not ethical – and it will cause major legal headaches.”

Krzeptowska agreed: “We’re improvising and experimenting constantly,” she said, describing her studio’s work as a wild mix of artistry, hands-on manual work and emerging tech.

This hybrid, cross-disciplinary approach sits at the core of the International Screen Institute’s mission. During the talk, chairwoman Andrea Elisa Pisu presented the institute’s range of training offers for audiovisual professionals, underlining the importance of preparing producers, legal experts, creatives, film commissioners, talent agencies and financial controllers alike.

The panellists collectively advocated for a proactive, informed use of AI. Lund Thomsen reminded the audience that even voice data are subject to personal rights legislation in some countries. “Have a conversation,” she said. “Get consent for the material you use. This isn’t just about copyright; it’s about personal data, too.”

Puhm, who admitted she’s “a terrible drawer but always had a clear visual mind”, said she now uses text prompts to generate images for her teams. “It’s liberating. It doesn’t replace a production designer, but it speeds up the process. You can reiterate your vision faster. That’s huge.”

The session closed with a strong consensus: AI is no longer a speculative horizon; it’s a working reality. Professionals must understand how to use these tools ethically and creatively while preparing for the rapid transformations to come.