Adam Shankman Denies AI VFX Use in 'Stop! That! Train!'

Share this post:
Adam Shankman Denies AI VFX Use in 'Stop! That! Train!'
Director Adam Shankman issued a public statement on June 2, 2026 calling social media claims that his RuPaul Charles comedy used generative AI in its visual effects "patently not true". The film opens June 12 via Neon. It was accused of AI use based on a critic's Letterboxd review and the presence of an AI company in its end credits.
Where the Accusation Started
Filmmaker and VFX artist Gloria Cook posted a Letterboxd review on May 28, 2026 following an advance screening of "Stop! That! Train!" Cook wrote that the film featured "one of the most conspicuous uses of AI I've seen in a film, with a lot of VFX looking like gen AI". She also flagged Acme AI, listed as a VFX and AI partner in the end credits.
Cook's professional background in VFX and filmmaking gave the claim traction. The review spread widely online before the film's opening.
Shankman's Statement
Shankman responded on Instagram the same day the story broke. "This is patently not true. Every shot in Stop! That! Train! was made by human hands. There are a sum total of ZERO shots conceived by AI in the movie".
He went further: "We employed hundreds of VFX artists who all killed themselves getting this out for release and not one job was taken out of human hands". Shankman described the project as "Fully Human Made" and called on audiences to "get out there and support all these magnificent talented and PAID artists" when the film opens.
What Acme AI's Role Actually Was
A source familiar with the production told Variety that Acme AI was contracted exclusively for visual effects work, and that any AI use was confined to background workflow processes and did not appear on screen. The company's presence in the end titles, standard for production service vendors, was the sole trigger for the public accusation.
That distinction, between an AI vendor supporting your pipeline and generative AI appearing in your frames, has no formal definition in any widely adopted credit standard. That ambiguity is what the controversy hinged on.
The New Standard of Disclosure Hollywood Needs
The Shankman case is an inversion of the pattern studios have maintained for years. Productions that used AI for voice cloning, background VFX, or title sequences typically disclosed it only after journalists asked, or not at all. Shankman was in the opposite position: defending human work against an accusation of AI use that he says was never there. But the public had no way to verify either claim from the credits alone.
That is precisely the gap the Human Provenance in Film standard, launched at the Cannes Film Market in May 2026, was built to close. The initiative defines three tiers of AI involvement: No AI Used, Assistive AI, and Generative AI. A production using an AI vendor for background VFX pipeline work would fall under Tier 2, Assistive AI, not Tier 3. Had that label traveled with "Stop! That! Train!" through distribution and marketing, Cook's review would have had accurate context from the first frame, and Shankman would not have needed to defend his crew on Instagram.
The underlying technical problem is harder to solve. Current AI detection methods are unreliable enough that audiences cannot use them to make credible judgments on their own. The C2PA Content Credentials standard would embed cryptographic provenance data at the shot level, recording exactly which tools touched each frame. But C2PA is not yet standard in theatrical production workflows. Until it is, end credits remain the only public signal, and as this case shows, a vendor credit and an on-screen credit look identical on paper.
Shankman's statement is unusually direct by industry standards. Most directors facing AI accusations have answered more obliquely, or left their publicists to handle it. That he named the number of VFX artists, described the film explicitly as human made, and tied his defense to their labor signals how much the accountability expectation has shifted. Directors now defend their crews before the film opens.
Productions that generate AI content as part of their creative workflow can work with the latest text-to-video and image-to-video models through AI FILMS Studio.
Sources
Variety | The Hollywood Reporter | Deadline | The Wrap | AV Club | BroadwayWorld
Continue Reading
Video & LipSync
- Video Generator
- Text to Video
- Image to Video
- Start-End Frame to Video
- Draw to Video
- Motion Control
- Video Enhancer
- Video Upscaler
- Video to Video LipSync
- Audio to Video LipSync
- Image to Video LipSync
- Video FaceSwap
- Seedance 2
- OpenAI Sora 2
- Kling 3.0
- Kling O1
- Google Veo 3.1
- LTX 2.3
- Kling O1
- Hailuo AI
- Luma Ray
- Kling 3.0 Motion
- Topaz Upscaler
- InfiniteTalk Face Swap
Image & Edit
- AI Character
- AI Actor
- Art Generator
- Text to Image
- Image to Image
- Draw to Edit
- Image Training
- Remove Background
- Image Enhancer
- MidJourney 8.0
- OpenAI GPT Image 2.0
- Kling Image 3.0
- NanoBanana Pro
- Minimax Image
- NanoBanana 2
- Kling Omni 3
- FLUX 2
- WAN 2.6
- Z-Image
- SeedEdit 3.0
- GLM-Image
- Omnigen 2
- Seedream 4.5
- Background Erase Network 2 (BEN2)
