Seattle Times and Newsday sue OpenAI and Microsoft over AI training

This lawsuit is another sign that the fight over training data is moving from angry op-eds to concrete legal risk. The Seattle Times and Newsday are not global giants, but they are influential regional papers that have taken money from both Microsoft and OpenAI in the past. Suing former partners signals that goodwill and fellowship payments are no longer enough to offset the perceived threat from generative models trained on their archives.

For the AGI race, the strategic issue is not this single case, but the cumulative effect of dozens of similar suits. If courts start treating news content differently from generic web text, frontier labs may need to carve out entire categories of material from training sets or pay to license them. That pushes the industry toward a more closed, capital-intensive model where only a handful of firms can afford fully compliant datasets, reinforcing the moat for incumbents like OpenAI and Microsoft.

At the same time, clearer case law could ultimately stabilize the environment. If the courts define what “fair use” looks like for training, labs and data providers can structure deals instead of litigating every edge case. In the short term, though, this adds legal drag and reputational pressure to companies already under scrutiny for safety and transparency.