CBC News analysis finds thousands of Canadian authors, books in controversial dataset used to train AI
A CBC News investigation has found at least 2,500 copyrighted books written by more than 1,200 Canadian and Québécois authors were shared online as part of a massive — and now defunct — dataset used to train artificial intelligence.
A CBC News investigation has found at least 2,500 copyrighted books written by more than 1,200 Canadian and Québécois authors were shared online as part of a massive — and now defunct — dataset used to train artificial intelligence.
