Jaron Lanier, a pioneer of virtual reality and Microsoft researcher, has long criticized the business models of major technology platforms. He argues that companies profit from user data and free labor while failing to compensate the creators whose content trains artificial intelligence models.
Lanier proposes a 'data dignity' model where corporations pay royalties to users for their digital contributions. He views AI not as an autonomous entity, but as a collective product of human effort. He suggests that the 'black box' of AI is actually filled with human data and labor.
The ongoing legal landscape reflects these tensions. On July 20, 2026, a federal court in Northern California approved a $1.5 billion settlement in the Bartz v. Anthropic case. This agreement covers 482,460 works used to train AI models without authorization.
Under the settlement, payments average approximately $3,000 per work before expenses. Publishers like Bloomsbury, which had 14,087 titles included in the settlement, expect to receive around $18.7 million. The first payments are scheduled to be distributed by November 15.
