Director Steven Soderbergh used generative AI to restore and animate archival footage in his documentary "John Lennon: The Last Interview" at the Cannes Film Festival [1].
The project has ignited a broader debate over the ethics of AI-generated content in non-fiction storytelling. Critics argue that altering historical records through synthetic media may undermine the fundamental authenticity required for the documentary form.
The film premiered during the festival, which ran from May 12 to 23, 2026 [2]. Soderbergh utilized the technology to enhance footage from Lennon's final interview, conducted on Dec. 8, 1980 [3].
"I owe a lot to Meta for the AI tools we used – they let us bring the interview to life in a way that would have been impossible otherwise," Soderbergh said [4].
The filmmaker acknowledged that the technical choices were not universally accepted. "The AI work sparked a lot of debate at Cannes, with some critics saying it undermines the documentary form," Soderbergh said [5].
While some reports describe the reaction as an uproar, other perspectives suggest the technology was not the only issue. One film critic noted that the film would have struggled to find relevance even without the AI, suggesting the technology only amplified existing conversations about the work [6].
Soderbergh's approach involves using AI to bridge the gap between static archival materials and a cinematic experience. This method allows for a more fluid visual narrative, though it risks blurring the line between historical fact and digital reconstruction.
“"I owe a lot to Meta for the AI tools we used," Soderbergh said.”
The controversy surrounding Soderbergh's documentary highlights a growing tension in the film industry between technological innovation and journalistic integrity. As generative AI becomes more capable of recreating human likenesses and speech, the industry must establish new standards for disclosure and authenticity to prevent the erosion of trust in historical archives.
