This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A filmmaker used ByteDance's Seedance 2.0 tool to generate a viral, realistic video depicting celebrities Brad Pitt and Tom Cruise in a fight. The video sparked significant backlash from Hollywood studios, the Motion Picture Association, and labor unions like SAG-AFTRA, who cited unauthorized use of likeness and copyrighted material. Disney issued a cease-and-desist letter to ByteDance, which subsequently stated it would strengthen safeguards and temporarily paused the ability for users to upload images of real people.
A filmmaker reportedly used ByteDance's AI video tool Seedance 2.0 to create and post a purportedly realistic clip depicting Tom Cruise fighting Brad Pitt, which then circulated widely online. Industry groups and studios publicly alleged the tool enables unauthorized use of copyrighted material and performers' likeness, and at least one studio reportedly sent a cease-and-desist letter. ByteDance said it respects IP and would strengthen safeguards.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems capable of creating economic or cultural value, including through reproduction of human innovation or creativity (e.g., art, music, writing, coding, invention), destabilizing economic and social systems that rely on human effort. The ubiquity of AI-generated content may lead to reduced appreciation for human skills, disruption of creative and knowledge-based industries, and homogenization of cultural experiences.
Also classified as
The video generated highly realistic but entirely fabricated dialogue about Jeffrey Epstein and Russian operations attributed to real people.
The incident highlights Hollywood's deep concern that highly realistic AI-generated videos will devalue human creative work and lead to widespread job losses.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was initiated by a human filmmaker who intentionally prompted the AI tool to generate the video, which occurred after the tool was publicly deployed.
Added categories peak at 2: Minor (Psychological Harm, Direct)
An Irish filmmaker used ByteDance's generative video tool to create a highly realistic, viral deepfake of US celebrities. While causing significant economic and intellectual property concerns within the entertainment industry, the incident has minor national security implications, representing an evolved capability in synthetic media generation rather than a coordinated information warfare campaign.
The report describes Seedance 2.0 as a 'generative video tool' that operates on direct user prompts ('This was a 2 line prompt') to produce deterministic video outputs without independent planning or autonomous decision-making.
Limited Risk: The system is used to generate AI-generated content such as deepfakes, which requires transparency obligations under the Act. The report describes Seedance 2.0 generating 'AI-generated versions' and 'lifelike versions' of real actors.
The system was used to generate highly realistic, synthetic video and voice representations of real celebrities without their consent.
The reports provide consistent, detailed accounts of the creation of the viral video, the tool used (Seedance 2.0), the developer (ByteDance), and the specific reactions from Hollywood stakeholders. There is little ambiguity regarding the facts of the incident.
The incident could be viewed purely as a copyright dispute rather than an AI safety issue, but the systemic implications for generative AI and creative labor align it with AI safety taxonomies.
Technology
High Realism of Seedance 2.0
Allows creation of lifelike deepfakes indistinguishable from real footage.
Unrestricted Image Uploads
Allowed users to upload images of real people to generate custom content.
Advanced Motion Handling
Enabled realistic generation of complex physical interactions like fighting.
Data Inputs
Unauthorized Use of IP
Alleged training using a pirated library of copyrighted Disney characters.
Likeness Data Without Consent
Use of real actors' faces and voices without their permission.
Human Factors
User Intent to Create Deepfakes
Users generated controversial videos of celebrities as jokes or tests.
Fear of Job Displacement
Industry professionals fear AI tools will replace human actors and writers.
Lack of Consent for Likeness
Actors had no control over how their likenesses were deployed by users.
Process and Methods
Inadequate Upload Safeguards
Lack of initial filters to block real person image uploads on launch.
Weak IP Infringement Checks
No robust mechanisms to prevent generating copyrighted characters.
Easy Prompt-Based Generation
Two-line prompts allowed rapid generation without oversight.
Management
Prioritizing Launch Over Safety
Released Seedance 2.0 without sufficient safeguards against IP theft.
Inadequate Risk Assessment
Failure to foresee misuse of celebrity likenesses and copyrighted assets.
Regulatory Environment
Lagging Copyright Regulations
Lack of clear legal guardrails for AI-generated likenesses and IP.
Lack of Pre-release Oversight
No regulatory review required before deploying powerful generative tools.
