Why Would Meta Download So Much Porn? (www.theatlantic.com)

Meta is facing legal action from Strike 3, the parent company of Vixen Media Group, for allegedly downloading nearly 3,000 of its copyrighted adult videos and other illicit materials, including celebrity nudes and hacking tools, over the past two years. The case illustrates the growing scrutiny on tech companies’ data-gathering practices for AI training, raising concerns about mass piracy and ethical boundaries in the industry. Strike 3 claims that their evidence reveals Meta not only downloaded this content but also made some of it publicly available, complicating Meta's defense that such downloads were solely for personal use by employees. The implications of this case extend beyond copyright infringement; they highlight the ethical dilemmas faced by AI/ML companies in data acquisition. While Meta argues against the use of this content for AI purposes, the existence of such files could facilitate the development of algorithms that identify “unsafe” content or even aid in the optimization of AI models. Additionally, with BitTorrent being a common method of mass downloading, this situation reflects a broader trend within the AI community, where many companies, including OpenAI and Anthropic, also rely on similar practices. As the industry navigates its efficiency and compliance concerns, the outcome of this lawsuit could set significant precedents regarding data sourcing methodologies in AI development.