Cannes, AI and the New Auteur: What Generative AI Means for Studios and Festivals

At Cannes, AI-generated films forced studios, festivals and creators to reconcile three things at once: rapid economics, contested authorship and rising legal risk. Executives need a practical playbook—how to pilot the technology without handing the courts and audiences the bill.

• Executive summary
  • Generative AI is already producing AI-generated films at scale: submissions to the World AI Film Festival (WAIFF) jumped from ~1,000 to ~5,000 in one year.
  • Studios see economics: lower costs and the ability to produce more mid-budget projects. Filmmakers and festivals worry about authorship, emotional depth and copyright.
  • Posthumous synthetic performances and works resembling existing characters exposed legal and reputational fault lines.
  • Five immediate actions for leaders: run pilots, update contracts, create governance, reskill staff and adopt disclosure policies.

What happened at Cannes

The World AI Film Festival (WAIFF) brought an unmistakable techno-cultural clash to Cannes: screenings of AI-generated cinema, investor interest from high-profile filmmakers and heated public reactions from veteran directors. WAIFF’s submissions surged to roughly 5,000 entries, up from about 1,000 the year before, signaling rapid adoption of generative AI tools for image, animation and VFX work.

Organizers at Cannes’ main festival drew a clear line: AI-made films would not be eligible for the Palme d’Or. As the festival put it:

“AI imitates very well but it will never feel deep emotions.”

That position framed the cultural debate—can algorithmic processes produce the kinds of lived, intentional work festivals celebrate? The WAIFF screenings made the question concrete. Some shorts were startlingly effective; others were uncanny or derivative. One entry removed from consideration resembled Aardman’s Wallace and Gromit so closely it raised copyright alarms. Director Mathieu Kassovitz reacted viscerally on social media—an eruption that captured how raw the debate has become.

The new economics of AI in film production

Generative AI is reshaping budgets and decision cycles. Joanna Popper, a film & tech executive and WAIFF judge, summed the studio rationale succinctly: AI lets producers have “more shots on goal” by making several ~$50M AI or hybrid films instead of placing all their chips on a single $200M production. The math is simple: faster iteration, cheaper VFX and the ability to A/B test concepts at scale.

Real examples illustrate the gap. Filmmaker Dario Cirrincione paid about €500 for an AI-generated VFX sequence that would traditionally cost nearer €20,000. That delta changes vendor relationships, hiring models and portfolio strategies. Instead of a one-off blockbuster, studios can field multiple mid-budget bets and optimize with audience data—bringing a Silicon Valley playbook into Hollywood.

Investors and creatives are moving accordingly. Names such as Ron Howard, James Cameron and Matthew McConaughey have backed AI film technology; major studios—including a public signal from Paramount—have stated AI will affect many areas of their business. That does not mean traditional filmmaking disappears, but it does mean production workflows and capital allocation will shift.

Creative, cultural and ethical fault lines

Technical capability does not automatically equal artistic legitimacy. Festival leaders and jury members expressed both curiosity and resistance. Agnès Jaoui, who chaired a jury, said accepting an AI role drew public backlash: “Ever since I accepted … everyone has been yelling at me. Are you validating AI?” Gong Li took a more exploratory tone: “AI can be controversial. But it can also open new ways to imagine stories. Let’s explore this together.”

Cannes’ president summarized the preservationist stance:

“A film is not an assembly of data; it is a personal vision.”

At the same time, veterans like Claude Lelouch embraced AI as a tool for creative freedom. The festival opening included an 80-piece orchestra performing Ravel’s Boléro—a deliberate reminder that live human artistry still holds symbolic and emotional weight. The result is a pragmatic split: some see augmentation and efficiency; others see dilution of authorial intent and a marketplace where imitation can outrun respect and compensation.

Legal landscape: copyright, likeness and consent

Generative models are trained on vast corpuses of human-created work. That fact is the root of multiple legal and ethical questions: who owns the inputs, who controls derivative outputs, and when does augmentation become infringement? WAIFF’s Wallace-like short and an AI-generated trailer invoking Val Kilmer’s likeness made those flashpoints public.

Key legal arenas to watch:

• Copyright and training data: Models often ingest copyrighted material. Licensing frameworks and transparency around training data will be essential to reduce litigation risk.
• Personality and publicity rights: Synthetic resurrections or close facsimiles of living or deceased performers raise estate, consent and reputation issues.
• Advertising and disclosure: Audiences expect honesty about synthetic elements. Failure to disclose can create consumer backlash and regulatory scrutiny.

Regulatory efforts—like the EU AI Act and localized publicity-rights laws—will shape risk. Businesses should not wait for final rules; they should build interim policies that respect creators and reduce downstream exposure.

Business playbook: five actions for leaders

• Run pilots, not bets. Start with controlled AI-assisted shorts or VFX pilots to test costs, creative outcomes and workflows. Treat them as learning investments and measure time, cost and audience response.
• Update contracts and procurement. Require vendors and model providers to disclose training data sources, warranty IP clearance and include indemnities for third-party claims.
• Create an AI content governance board. Combine legal, creative, brand and product stakeholders to approve synthetic assets, set disclosure rules and pre-clear likeness use.
• Invest in reskilling. Fund upskilling for VFX, editorial and legal teams so they can integrate generative tools effectively—shifting roles rather than simply cutting headcount.
• Adopt audience disclosure policies. Decide when and how to label synthetic performances or trailers to preserve trust—transparency reduces surprise-driven backlash.

Key questions leaders are asking (and practical answers)

• Who owns the works used to train AI models?
  

  Ownership is contested. Until clear licensing frameworks exist, require provenance disclosures from vendors and consider negotiated payments or revenue shares for material contributors whose work materially influences outputs.

• Can AI-generated performances capture deep human emotion?
  

  AI can mimic timing and expression, but many filmmakers argue it lacks lived intentionality. Use AI where it augments—not replaces—human direction, and test for audience resonance before scaling.

• Will studios pivot to portfolios of lower-cost AI/hybrid films?
  

  Evidence suggests a shift toward portfolio experimentation. The expected outcome: more mid-budget attempts, faster iteration and a different mix of staffing and vendor services.

• How should posthumous or synthetic performances be handled?
  

  Only with clear authorization from estates or rights holders—and with a rigorous ethics and disclosure protocol. Treat synthetic likenesses as a high-risk use case requiring express legal clearance.

• What should business leaders do next?
  

  Combine operational pilots with governance changes: test AI in limited productions, tighten contracts, form an approval board, fund reskilling, and adopt disclosure rules to protect brand and creators.

Generative AI is not a distant possibility. It is already changing production economics and creative workflows while exposing gaps in rights, consent and governance. The choice facing studios and festivals is not binary—preserve old models or abandon them—but practical: how to integrate AI tools to accelerate creativity and lower costs while protecting authors, performers and brand trust.

Start small, put governance first, and treat synthetic media as both an operational opportunity and a policy problem. That balance—practical experimentation plus rigorous safeguards—will determine who thrives in the next era of filmmaking.