According to media reports, multiple class-action lawsuits have been formally filed against Apple, Amazon, and OpenAI. The plaintiffs accuse these tech giants of bypassing YouTube's anti-scraping protection mechanisms to illegally download and use millions of video data to train their AI models.

Key Allegations: The "Plagiarism" Behind the Panda-70M Dataset

This lawsuit was initiated by three YouTube channels: Ted Entertainment, Matt Fisher, and Golfholics. The focus of the dispute is a dataset called Panda-70M:

Data Indexing: This dataset breaks down massive videos into tens of millions of independent training samples through URLs, video IDs, and timestamps.

Evasion Behavior: The plaintiffs point out that in order to extract these segments, research teams from companies such as Apple had to frequently access and intercept original video content. This process is accused of deliberately bypassing YouTube's copyright protection system established to protect creators.

Substantial Evidence: The Apple team mentioned using this dataset for training in their research paper on the video generation model STIV.

Lawsuit Demands: Request for Jury Trial and Compensation

The plaintiff represents all content creators in similar situations and has made several strict claims to the court:

Statutory Damages: According to U.S. copyright law, they demand monetary compensation at the maximum legal limit.

Injunctive Relief: They request that the defendants and their affiliated parties immediately stop infringing behavior and cease using copyrighted content to train models.

Fair Review: They apply for interest before and after the judgment, and require the defendant to bear the lawyer's fees and litigation costs.

Industry Background: The "Tragedy of the Commons" in the AI Era

This is not the first time AI giants have been involved in controversies over training data. As the demand for high-quality video data increases for large models, finding a balance between protecting creators' rights and promoting technological advancement has become a key challenge in global tech regulation.

Pressure on OpenAI: After Musk's antitrust lawsuit, OpenAI faces further moral questioning about the legality of its data sources.

Apple's Transformation Challenges: Apple, which has always emphasized privacy and copyright protection, is accused of "intentionally evading protection systems" in this case, which poses a significant blow to its brand image.

Extended Developments: The Intensifying Battle for Talent

Outside of legal disputes, the rivalry among giants has already escalated. According to reports, Apple is trying to prevent its core engineers from moving to OpenAI by offering high bonuses. Meanwhile, OpenAI continues to recruit talent from Apple's hardware team, aiming to lay the groundwork for future AI hardware.

Conclusion: A Collective Counterattack by Original Creators

As AI-generated videos become increasingly realistic, the ownership of their underlying "raw materials" has become an unavoidable legal boundary. The outcome of this class-action lawsuit may set an important precedent for the legal use of training data in AI globally.