Multiple class-action lawsuits have been filed against Apple, Amazon, and OpenAI. The plaintiffs, three YouTube channels, allege these tech giants illegally bypassed YouTube's anti-scraping protections to download and use millions of videos for AI model training, focusing on the controversial Panda-70M dataset. The core accusation centers on the systematic indexing of videos into tens of millions of clips using URLs, video IDs, and timestamps. Plaintiffs argue this process intentionally circumvented YouTube's copyright protection systems. Evidence includes Apple's research paper for its video model STiV, which cited using this dataset. The lawsuits seek statutory damages under U.S. copyright law at the maximum allowable rate and injunctive relief to halt the use of copyrighted content for training. They also demand pre- and post-judgment interest, along with coverage of legal fees. This case highlights the ongoing tension in the AI industry between the demand for high-quality training data and creator rights. It follows previous legal challenges for OpenAI regarding data sourcing and poses a significant brand risk for Apple, known for its privacy and copyright stance. Simultaneously, a fierce talent war is intensifying, with Apple offering bonuses to retain engineers targeted by OpenAI, which is actively recruiting from Apple's hardware team to build its AI hardware capabilities. The outcome of this collective action by creators could set a crucial global precedent for defining the legal boundaries of using data to train AI systems.