By the sounds of it, Anthropic is jealously guarding the information about what datasets it used to train its AI model like a crown jewel. Bartz book authors just filed a letter objecting to the alleged restriction that Anthropic is now imposing: “Anthropic claims Plaintiffs can only view this evidence in a highly-monitored and restrictive inspection environment.”
Will be interesting to hear Judge Alsup’s view on this controversy. Unlikely to be pleased.
