
Industry Talk: End times – AI and documentary
AI and end times have become the conversation of the moment, stirring the same unease. What does AI actually offer documentary storytelling as we engage with these narratives? And what is at stake?
The Industry Talk: End times – AI and documentary on November 18 weighed in on these burning questions, examining the creative and political implications of AI use in film. Moderated by Kat Cizek (Artistic Director of Co-Creation Studio at the MIT Open Documentary Lab), the discussion was joined by Daniel Mann (The Recce (2026)), Kyle McDonald (Apocalypse Early Warning System (2026)), and Isabel Joffily and Rita Toledo (Morro Grande (2026)), exploring how they used AI in their work to grapple with oppressive systems and apocalyptic narratives themselves sustained by this technology. This year, they were among the filmmakers who declared their use of AI when applying to IDFA, a disclosure the festival requested for the first time. Michaela Pnacek(ova) (XR creator and researcher at the Norwegian University of Science and Technology) also took part in the discussion, introducing a Horizon Europe project, European AI Commons – Shaping AI for the Creative and Cultural Industries (EAiCCI), launched in September 2026 to advance transparent and ethical AI that serves the public good.
Asked about his use of AI in The Recce (Paradocs), Mann observed the “ambiguity” it created, which proved creatively fertile. The Recce centers on a 1904 expedition to Uganda sent by European leaders of the Zionist Organization to survey the then British protectorate as “a potential Jewish state.” Like other such expeditions, it assessed the land and its useful botany and resources for settlement and extraction, “a core tenet of settler colonialism,” the filmmaker noted, and it was “a rehearsal of a movement that adopted colonialism wholeheartedly.” The film is composed of present-day location scouting, a form usually marked by establishing shots and empty frames, which echoes in some ways the approach of appraising the land as a test site for settlement.
AI was then deployed to visualize the expedition’s field notes, Mann explained, producing “absurd images” that speak to how “the landscape was imagined as fantasy” that projected the colonizer’s desires back onto it. This produced a tension between the photographs, whose visual evidence served to promote settlement, and the AI’s “sanitized” images, which created a “purified” world that cleared the frame of its inhabitants while amplifying the apparent lushness of its flora. “I found it incredibly disturbing, but it captured the [apparatus] that enabled the violence perpetrated through settlement itself,” he added.
The Recce “linked AI to systems of power” and “historicized the relationship between the images and colonialism in a very profound way, Cizek reflected. She then turned to the directing duo, Joffily and Toledo, who used AI in Morro Grande (International Competition) to shed light on a story affected by “chemical colonialism.” Citing the Brazilian Indigenous author Ailton Krenak’s remarks on the end of the world, Toledo reframed apocalypse not as a future event but as a slow “historical process” of environmental and social destruction that Indigenous peoples have long endured. In this vein, the filmmakers engaged with AI to create “a different past” for their characters, which “could inspire a different future.” Drawing on archives, Joffily and Toledo also worked with Midjourney to generate photographic fabulations of a story of a man (Vaguinho), who took his own life by drinking pesticide. They also used this method to depict a ‘90s Brazilian rural punk movement that resisted agrochemical framing. Asked whether their use of AI renegotiated the contract with audiences, they noted the practice was part of fabulation that blurred the line between memory and imagination, featuring the narrator’s “retelling and remembering” of his friend Vaguinho’s story, itself an inherently unreliable form of testimony.
The Talk then moved to artist-coder McDonald behind Apocalypse Early Warning System (IDFA DocLab Competition for Digital Storytelling), who created a real-time dashboard and art installation that monitors global private jet movements to spot unusual spikes that might signal a looming disaster. Unlike his co-panellists, McDonald approached AI as “a tool where the truth is defined by data.” The project partly grew from his distrust of news shaping our picture of the world. He noted that insider trading on oil futures, coups, and other major events cause concern and are a reason for the public to start relying on data.
The discussion concluded with a Q&A that touched on several important questions. While AI may democratize filmmaking amid shrinking budgets, its use needs caution, as it has “a long list of risks” and might introduce a new kind of alienation into the creative process, which “comes at a huge price,” Mann said. Joffily and Toledo agreed, adding that AI is not the first technology to be met with suspicion, following drones, first designed for military use, and cameras, which can’t be naively approached as bearers of truth, as they are also rooted in extraction and exploitation. However, as Cizek noted, AI can serve as a gesture of counter-surveillance and counter-colonialism, if used thoughtfully. To that end, researcher Pnacek(ova) presented the EAiCCI project, which seeks to map the realities and (future) practices of AI in Europe, and to foster conditions in which artists can maintain agency and creative control when collaborating with the technology. This interdisciplinary effort, involving researchers, creators, technology producers, and policymakers, also aims to develop checklists and best-practice guidelines, and to codesign AI tools based on transparent, value-aligned models.
