The Seattle Times and Newsday are suing OpenAI and Microsoft, adding two more names to an already growing legal fight over how the companies get their training data.
The two papers jointly allege that OpenAI and Microsoft methodically scraped their articles in a way that bypassed paywalls, rather than accessing that content the way a paying subscriber would. That's a specific, technical claim, and the EU's general-purpose AI code of practice draws the same line: the voluntary framework commits signatories to not circumvent access restrictions, and it names subscription models and paywalls specifically. The Seattle Times and Newsday now join other newsrooms already pursuing OpenAI and Microsoft in court over how the companies obtained their content.
That framing matters. Instead of relitigating whether training on published text counts as fair use, the newsrooms are pointing to a narrower, more testable action: circumventing a technical barrier a reader has to pay to cross. That's a tighter claim for a court to evaluate than the broad fair-use arguments running through other AI copyright suits.
The EU's code is voluntary and European. Whether bypassing a paywall is illegal in the US is a question for an American court, not a Brussels framework. That's exactly what this lawsuit is set up to test.