Notes from the AI & Film Academy, a five-day program by Swissnex San Francisco and the Locarno Film Festival.

In early April, Swissnex in San Francisco hosted four filmmakers and creative technologists for the AI & Film Academy, an inaugural five-day program designed jointly with the Locarno Film Festival‘s established Academy program. The premise was straightforward: bring practitioners working at the edge of cinema and AI into proximity with the companies building the tools, and see what happens when creativity and curiosity meet infrastructure.

The conversation that emerged resists the binaries that dominate public discourse on AI. It was not a debate about utopia or collapse, replacement or liberation. It was something more useful: a working session among emerging filmmakers and artists already inside the systems, asking what it means to shape technology rather than be shaped by it.

Swissnex in San Francisco’s Creative Industries Manager Lucas Hagin co-organized the Academy and shared his insights across four themes that emerged in conversation.

San Francisco – April 23, 2026

By Lucas Hagin, Creative Industries Program Manager at Swissnex in San Francisco

At Swissnex in San Francisco, we are actively involved in the conversation around media and AI, and how generative AI is currently forcing every creative industry to revisit its purpose and its future. The week was an opportunity not just for the Academy participants, but for me to explore topics I am deeply invested in. What I found was a forum where practitioners were already past the binary of pure enthusiasm or skepticism. They were not just exploring the theory of AI but sharing their own workflows, workarounds, and convictions. Strong opinions were met with pushback, nuanced discussion, and patient arguments.

The Big Picture

Evelyn Kreutzer, PhD, a German post-doctoral researcher, video essayist, and filmmaker at Università della Svizzera italiana in Lugano, Switzerland, named one obvious problem most directly:

Recently, across screenings, visits, workshops, panels and conversations with leading art institutions, companies working on tools for media and entertainment, AI film studios , researchers, communities, and local filmmakers, participants encountered AI in very different settings and use cases. They see it as a layered set of tools, contracts, infrastructures, and, crucially, people. What the week did, above all, was bring people into the same room and make their positions legible. As Evelyn also noted:

Kevin B Lee, a native of the Bay Area and Professor for the Future of Cinema and Audiovisual Arts at Università della Svizzera italiana, reframed the stakes from another angle. The conversation around tools and creative agency, he noted, assumes a degree of continuity that doesn’t yet exist. Platforms appear and disappear. Workflows built around one model become obsolete when that model is discontinued or acquired. The relationships between artists and tools feel, in his words, “very indiscriminate, very promiscuous, and very unloyal.” In his eyes, he sees the current playing field of generative AI as:

His deeper question was whether the current volatility will ever settle into something stable enough to support sustained artistic practices. Film festivals, residencies, university programs, and initiatives like the Academy itself depend on continuity. So does any creative work that requires more than a single production cycle to develop. Without infrastructure that holds still long enough for creatives to build on, the expanded possibilities risk remaining permanently provisional. Against that backdrop, the question is not only what generative AI will do to creative work, but whether platforms and creators will find solid enough footing to build on longterm.

Scaling Imagination

If one register of the week was sober, another was openly enthusiastic, Andrea Gatopoulos, an Italian and Greek film director, producer and distributor, framed the shift as an expansion of what a filmmaker can afford to imagine:

For Andrea, this expansion is felt as a newfound creative agency. Cinema, in his telling, is “a continuation of my childhood playing, a very serious game.” The new tools interest him precisely because they restore tempo and tactility to a process that had grown slow and procedural. The risk of long production cycles, he noted, is that “you make the movie based on this past urgency that then slowly becomes real.” Compressing the feedback loop is, in this sense, an aesthetic question as much as an economic one.

Breaking the Frame

Much of the discourse around AI operates at the level of industry forecasts and hypothetical futures. Rachel Maclean, a Scottish multimedia artist and filmmaker, brought the conversation back to the workshop floor. Her practice has long involved production with green screens, building fantastical worlds by layering performers against digital backdrops. For a decade, that method defined not just how she shot but what she could write. Scripts were shaped around the technology’s constraints: limited camera movement, no complex lenses, hours lost to technical requirements like removing shadows from the studio floor.

Her latest project changed that. By shooting conventionally and then taking the footage into AI-assisted post-production, Rachel found herself working without the familiar restrictions for the first time.

The liberation she describes is not a projection about what AI might one day enable but a concrete account of what shifted when one practitioner, deeply familiar with her medium’s limits, found a tool that removed some of them. Artistic choices that her previous workflow had made impossible were suddenly available. What stays with me, though, is a question: AI removed those constraints, but are we simply trading one set of limitations for another, now dictated by the mechanisms and workflows of generative models? There is a pattern emerging where we begin, sometimes without noticing, to follow the logic of AI: structuring prompts the way the system rewards, iterating toward outputs the model handles well, adapting our process to the tool’s preferences.

Productive Friction

Mayte Gómez Molina, a Spanish writer, researcher, and new media artist teaching at University of Applied Sciences Northwestern Switzerland, returned the conversation to craft. What interested her was not the elimination of work but its redistribution. Using generative AI, in her account, requires the same structural literacy as any other medium: knowing the tool’s grain, its failure modes, the points where intervention is possible. She also raised the question of constraint as a creative resource.

Mayte also pushed the group toward a register often missing from AI discourse: access. “In film or technology, where the tools are so expensive and the entry level is so expensive, this allows for people to have a try.” The digital output, she noted, does not have to be the end of the process. It can be a middle step, a scan before a cast, a mockup before a pitch, a bridge between an idea and the resources to realize it.

Unscripted

I came away from the week recognizing that some of the most interesting moments were when someone’s position visibly shifted. Linn Spitz, a Swiss interdisciplinary digital artist with a background in computer science, pushed the conversation in a different direction. Across five days of discussions, she noticed a pattern: certain creative tasks are clearly off limits to AI, even for many of the most enthusiastic adopters.

Her observation reframes the AI debate as a question of social norms rather than purely technical capability: Most people currently agree that ChatGPT scripts are still bad. They tend to produce a statistical middle ground: structurally competent, tonally generic, missing the specific voice or vision that makes a script worth producing. The question is whether this will remain the case. The lines being drawn now, often unconsciously, will harden into the working conventions of the next decade.

Linn was also candid about her own position shifting across the week. “I still hate big tech and the economic structures around it,” she noted, but added that being in conversation with her fellow participants had changed something. “I felt invited to question what my specific motivations are. My opinion toward how to use AI in art has shifted a bit more to the accepting, definitely more because of you guys than the tech representatives.” Part of what I appreciate about our work at Swissnex is witnessing these shifts in mindset.

To be continued…

What emerged across five days was not consensus but the creation of a working vocabulary. Liberation and limitation, play and constraint, access and authorship, frontier models and precision tools: these were not polar opposites but coordinates that practitioners move between, deliberately, as part of their craft. As Mayte put it at one point in the week, “things that are happening are in human nature; they’re just amplified by technology. Technology is not causing all problems.” The question, then, is how cultural practitioners hold their position inside that amplification.

At the close of the week, Evelyn noted that meeting these technologists gave her some hope. There are real human beings who are approachable, she said, working across a diverse spectrum of positions on AI, several of them actively curious about what artists and critical thinkers had to say.

The ambition of the AI & Film Academy was never to resolve AI’s place in cinema, but to bring creatives, scholars, and toolmakers into the same room long enough for the exchange to move past the noise. Participants left with sharper questions and, in some cases, small but meaningful shifts in their positions. They also left with a clearer sense of what their own agency looks like inside these systems: not as users of fixed products, but as practitioners whose choices, refusals, and experiments will help author the conventions still being written.

I see this as part of the red thread running through Swissnex’s 2026 Creative Industries program. The question is not whether artists will engage with AI. They already are. The question is whether spaces exist for them to do so as authors, with the time, peers, and proximity to make their decisions count.

The conversation continues this summer at the Locarno Film Festival, where the Academy’s participants will reconvene alongside a broader community of filmmakers, scholars, and industry voices.

The inaugural AI & Film Academy was hosted at Swissnex San Francisco in partnership with the Locarno Film Festival. Participants: Andrea Gatopoulos, Linn Spitz, Mayte Gómez Molina, Rachel Maclean. Co-organized by Evelyn Kreutzer, Kevin B. Lee, and Stefano Knuchel. The conversation continues at the Locarno Film Festival, August 5–15, 2026.

Lucas Hagin is the Creative Industries Program Manager at Swissnex in San Francisco

This is part of a newsletter published on LinkedIn. Subscribe here to get alerted about new articles.