Harvard University and the Boston Public Library are releasing collections of digitized materials to tech companies for AI training, the AP reports. Harvard is providing nearly one million books in 254 languages from the past 600 years, in a dataset called Institutional Books 1.0. The BPL will give collections of old newspapers and government documents. “It is a prudent decision to start with public domain data because that’s less controversial right now than content that’s still under copyright,” Microsoft deputy general counsel Burton Davis said. The library material includes original sources, which is lacking in much of the AI companies’ …