Runway AI Accused of Illegally Using YouTube and Pirated Videos for Model Training

The tool’s training on unlicensed YouTube channels and pirated media constitutes a breach of intellectual property rights, meeting the definition of an AI Incident under violations of IP law.

The article describes an actual instance where an AI system’s development (training data collection) infringed on creators’ intellectual property rights by scraping videos without consent, causing legal claims and reputational damage. This meets the definition of an AI Incident (violation of intellectual property rights).

The article describes an AI system’s development phase—specifically its model training—relying on unauthorized, copyrighted material. This constitutes a direct violation of intellectual property rights (OECD harm category c) and represents an actual AI Incident rather than a mere hazard or background report.

This describes an AI system’s development practice that appears to have directly violated copyright law by scraping and using protected content without authorization. That is a realized harm—an infringement of intellectual property rights—so it qualifies as an AI incident.

The reported scraping and use of copyrighted videos from YouTube channels (Netflix, Disney, individual creators) and pirated sites for training an AI model, without paying fees or obtaining permission, directly violates intellectual property rights. As this harm has already occurred, it qualifies as an AI Incident (violation of intellectual property rights).

Runway’s training of its AI system on stolen and unlicensed copyrighted videos constitutes a direct violation of intellectual property rights. This breach of legal obligations in the development of an AI system qualifies as an AI Incident under the framework’s category of IP rights violations.

The article describes a specific AI system (Runway Gen-3) whose development involved improperly collecting and using copyrighted videos to train its model. This constitutes a realized harm—breach of intellectual property rights and violation of terms of service—so it meets the definition of an AI Incident.

The event describes an AI system’s development and use that directly involves unlicensed training on copyrighted works, constituting a violation of intellectual property rights—a harm category under AI Incidents.

The incident involves the unauthorized scraping and use of copyrighted video transcripts and content—an AI system (Runway’s Gen-3 Alpha) was trained on pirated data, constituting a violation of intellectual property rights. This meets the criteria for an AI Incident because it has directly led to a breach of legal obligations intended to protect creators’ IP.

The article explicitly states that Runway used a crawler to download videos from YouTube channels and piracy websites to train its AI model, which likely includes copyrighted content from major companies. This unauthorized use of copyrighted material for training violates intellectual property rights. The AI system's development (training) directly involves this infringement. Hence, the event meets the criteria for an AI Incident due to violation of intellectual property rights caused by the AI system's development process.

Runway's AI system was allegedly trained on copyrighted YouTube videos without authorization, which directly implicates violations of intellectual property rights. The harm is realized as creators have expressed backlash and lawsuits are underway, indicating legal and rights-based harm has occurred. The AI system's development and use are central to this harm, fulfilling the criteria for an AI Incident under the OECD framework.

The article explicitly mentions that Runway AI used publicly available YouTube videos, including copyrighted content from creators and major studios, to train its AI model. This use of copyrighted material without clear authorization constitutes a breach of intellectual property rights, which is one of the harms defined under AI Incidents. The involvement of AI in training models on such data is clear, and the harm (violation of rights) is directly linked to the AI system's development and use. The mention of legal actions against other companies for similar practices further supports the classification as an AI Incident rather than a hazard or complementary information.

The article explicitly mentions that Runway used Nintendo's content without permission to train their generative AI model, which is a direct violation of intellectual property rights. This harm has already occurred as the training was done without consent, and Nintendo is preparing legal action. The AI system's development process is directly linked to this harm, fulfilling the criteria for an AI Incident under violations of intellectual property rights.

The AI system (Runway's video generation tool) was developed using data obtained in violation of copyright and platform rules, which directly breaches intellectual property rights. The unauthorized use of copyrighted content for training the AI model is a clear breach of applicable law protecting intellectual property rights, fulfilling the criteria for an AI Incident under the framework.

The event involves the use of AI systems (text-to-video generative AI) trained on scraped YouTube content, which is an AI system development and use issue. The scraping and use of content without permission directly implicates violations of intellectual property rights and platform rules, which falls under harm category (c) "Violations of human rights or a breach of obligations under the applicable law intended to protect fundamental, labor, and intellectual property rights." Since the scraping and training have already taken place, this is a realized harm, not just a potential risk. Therefore, this qualifies as an AI Incident.

The AI system (Runway's Gen-3) was developed using data scraped without consent from copyrighted sources, which constitutes a violation of intellectual property rights. This unauthorized use of copyrighted content in training the AI system is a breach of applicable law protecting intellectual property rights, thus meeting the criteria for an AI Incident under category (c).

The event clearly involves an AI system (Runway AI) that was developed using unauthorized copyrighted content, which is a violation of intellectual property rights. This fits the definition of an AI Incident because it involves a breach of obligations under applicable law protecting intellectual property rights. However, since the article does not describe any realized harm beyond the violation itself (e.g., no reported damage, no misuse of the AI outputs causing harm), the incident is primarily about the rights violation. The uncertainty about legality does not negate the fact that the AI system's development involved unauthorized use of copyrighted material, which is a breach of rights. Therefore, this qualifies as an AI Incident rather than a hazard or complementary information.

The article reveals that Runway's AI video generation tool was trained on pirated and stolen copyrighted content from major companies without authorization. This use of copyrighted material in AI training is a breach of intellectual property rights, fulfilling the criteria for an AI Incident under violations of intellectual property rights. The harm is realized as the unauthorized use has already occurred, not merely a potential risk.