Welcome to The Synthesis, a monthly column exploring the intersection of Artificial Intelligence and documentary practice. Co-authors shirin anlen and Kat Cizek will lay out ten (or so) key takeaways that synthesize the latest intelligence on synthetic media and AI tools—alongside their implications for nonfiction mediamaking. Balancing ethical, labor, and creative concerns, they will engage Documentary readers with interviews, analysis, and case studies. The Synthesis is part of an ongoing collaboration between the Co-Creation Studio at MIT’s Open Doc Lab and WITNESS.
We have been remiss in not writing lately. Meanwhile, documentaries made with—or about—AI abound. Every major festival so far this year has featured at least one. As we approach our second anniversary of this column, The Synthesis, we write to catch you up on our recent learnings from this latest crop of AI documentary films, dearest gentle reader, to help save you time. We have watched many AI documentaries. Others we have tried to chase down but have only read about, admittedly, as they are hard to access due to distribution and legal issues, so there is that caveat.
From our perch here at The Synthesis, we are seeing trends emerge across the suspect ones. These are problematic tropes that suggest projects feed into the mythologies of the AI industry, rather than debunk them, contributing to the Empire of AI (to quote the title and concept of Karen Hao’s foundational 2025 book). We have been hesitant to commit a whole column to any project individually, so instead, we have compiled a list of common signs of troubled waters below.
Here they are: nine red flags to watch out for in AI documentaries. And please do believe us, we have seen each one of these appear more than once.
Beware of documentaries that:
1. Feature a former tech giant who now regrets everything.
2. Promise to take you somewhere you’ve “never gone before” using AI-generated imagery. Keep an eye out for its use for dream sequences, which somehow always end up looking like the same vaguely ethereal AI slop we encounter every day.
3. Have a script written “entirely by AI.”
4. Are mostly made by—and/or almost exclusively feature—white men who all somehow know each other, with every interviewee introduced as an “AI expert.”
5. Are motivated by existential musing of a first-time dad about the correlation between human and artificial intelligence (with B-roll of said dad and baby, laced with voiceover pondering the definition of “intelligence”).
6. Hint at a new form of access, through AI, to an amalgamated immortal wisdom of a beloved cultural figure’s collected body of work and writings
7. Rely on a complicated set-up that has subjects appear as both themselves and as fictionalized characters in a reality-TV game-like scenario designed to rehearse saving the world from AI.
8. Include, every five minutes or so, someone saying, “The genie is out of the bottle.” Bonus points if they opt instead to talk about “Pandora’s box.”
9. Somehow still ends with the trite and pat question of, “What makes us human?”
Obviously, this is not a comprehensive list.
There are further considerations in evaluating the value of an AI documentary: basic issues of consent, purpose of use, transparency, source training data, IP, privacy, material consequences for the environment and human dignity, and many others that we have raised in previous columns and will aspire to address in upcoming editions. But as this is a short, summer installment of The Synthesis, we offer this slim but painfully human summary of warning signs for films that are both confusing and confused. If we have missed any other red flags, do let us know. Meanwhile, we will keep searching for the genuinely thoughtful, inventive, and critical works emerging at the intersection of AI and documentary practice. They do exist!
See you in the fall with back-to-regular programming of The Synthesis as we enter year three.
