Nvidia accused of trying to cut a deal with Anna's Archive (www.tomshardware.com)

Nvidia is facing accusations of attempting to negotiate a payment for "high-speed access" to Anna's Archive, a controversial online repository known for hosting copyright-infringing materials. Internal documents leaked by TorrentFreak suggest that the Nvidia Data Strategy Team sought this access to enhance the training of its large language models (LLMs). This revelation comes amid a class action lawsuit where Nvidia is already under scrutiny for allegedly infringing copyright by using materials from the Books3 dataset, which includes works from pirated sources. The correspondence indicates that Nvidia's management approved this questionable plan swiftly, potentially undermining their defense based on “fair use” principles. This situation raises significant concerns about the ethical implications of AI training practices within major tech firms. The proposed access would have involved around 500TB of data, accentuating the AI industry’s insatiable demand for vast datasets, often without regard for copyright laws. Such revelations not only highlight the contentious nature of AI model training but also amplify the ongoing debate around intellectual property rights, particularly as numerous authors seek justice against perceived exploitation of their works. With growing investigations and lawsuits, the Nvidia case could set a precedent affecting how AI companies approach data acquisition in the future.