When the gripping romantic thriller Plain Clothes made its official splash at the Sundance Film Festival, it didn’t just walk away with critical acclaim and the Best Ensemble award—it also stood out as a triumph of ethical machine learning integration. Directed by Carmen Emmi and starring Tom Blyth alongside Russell Tovey, the film handles highly sensitive historical themes centered on undercover police sting operations targeting gay men in the 1970s.
To protect the identities of real, innocent individuals captured in the raw, archival footage of the era, the production required a delicate touch that traditional masking tools simply couldn’t achieve without damaging the film’s immersion. Step in NYFA Media Arts Chair and Cinematic AI faculty member, Matt Galuppo. Serving as a key production asset, Galuppo utilized advanced machine learning technologies to meticulously alter and anonymize the archival backgrounds.
By altering faces, features, and identifying markers through a controlled generative pipeline, the team successfully shielded real-world identities while completely preserving the grainy, paranoid verisimilitude of 1970s surveillance footage. This application proves that AI tools, when guided by human ethics, serve to protect people while elevating the historical authenticity of independent cinema.