The Lawsuit and Its Claims
The Seattle Times and Newsday filed suit against OpenAI and Microsoft on 4 September 2026, according to startupfortune.com, accusing the companies of scraping paywalled journalism without permission or payment to train ChatGPT, Copilot and Bing AI. TechCrunch reported the filing on 5 September 2026. The complaint names both companies over the same alleged pipeline: ChatGPT on the OpenAI side, Copilot and Bing AI on the Microsoft side.
The suit departs from most publisher litigation in its requested remedy. Per startupfortune.com, it goes further than comparable cases by demanding the destruction of training data and models, and it accuses OpenAI of trademark dilution. Deletion, if ordered, would strip the disputed material from the defendants' systems rather than price it after the fact. Whether a court would treat model destruction as a proportionate remedy is unresolved; no ruling exists in the record.
The filing's language frames the dispute as existential. It argues that, with the advent of AI, the journalism industry could become "broken beyond repair," and describes generative AI as "a snake eating its own tail" that could "destroy the very organizations" producing the work the models ingest, per techcrunch.com. The metaphors point at a substitution argument: models consume reporting, then compete with the outlets that produced it. That substitution claim is the economic core beneath the copyright allegations.
Where This Fits in the Publisher Litigation Wave
The two newspapers are the latest publications to sue OpenAI and Microsoft, per techcrunch.com, joining a growing list of publishers pressing the same core claim: journalism scraped without license or payment, then replicated by systems that compete with the source. pluang.com places the start of this litigation line in 2023, with The New York Times' suit against the same two companies. The defendant pair has stayed constant across the wave; the plaintiff roster keeps expanding.
What separates this filing, on the record available, is the destruction demand. The reporting does not detail the remedies sought in earlier publisher suits, so a direct comparison is not possible from the evidence; what startupfortune.com states is that this complaint goes further than most by seeking deletion of models built on the scraped data. That escalation shifts the stakes from damages calculations toward product survival, and it gives the defendants a concrete reason to prefer settlement over an adverse ruling.
The accumulation itself carries weight. Each new plaintiff widens the factual record on scraping practices and adds outlets with paywalled archives to the claimant pool; startupfortune.com's account of the complaint centers on paywalled journalism. That suggests the litigation is moving from isolated grievances toward a body of case law the labs cannot resolve filing by filing. The open question is consolidation — nothing in the record indicates the suits will be combined or coordinated.
What the Defendants Say and What Comes Next
Microsoft's response, per pluang.com: the company expressed surprise at the lawsuit but is open to discussions to resolve the dispute. That combination — surprise paired with willingness to negotiate — indicates Microsoft would rather reach a commercial resolution than litigate a precedent on model destruction. OpenAI's response is absent from the available reporting; the record is silent on whether it has commented or filed any defense.
No court schedule appears in the reporting. The procedural path for a copyright case of this shape — motions, discovery over training corpora, trial — is not documented in the evidence, which records the filing and Microsoft's stated openness and stops there. What the destruction demand guarantees if the case proceeds is discovery into exactly what was collected and how, since the plaintiffs would need to establish their content sits in the training corpus.
Two signals matter from here. First, whether Microsoft's openness to discussions produces a licensing arrangement, which would convert the dispute into a priced transaction and undercut the destruction remedy. Second, whether further publishers file with matching demands; a second destruction claim would suggest a strategy rather than an isolated escalation. Neither outcome appears in the record. The filing is days old as of the latest reporting, dated 5 and 6 September 2026 across the sources.
Sources
- techcrunch.com — Seattle Times and Newsday Sue OpenAI and Microsoft Over AI Training
- techbuzz.ai — Seattle Times, Newsday Sue OpenAI, Microsoft Over AI ... (2026-09-05)
- pluang.com — Seattle Times and Newsday sue OpenAI and Micros... (2026-09-05)
- startupfortune.com — Seattle Times and Newsday Sue OpenAI and Microsoft Over News Scraping (2026-09-05)
- New Horizon daily digest, 2026-09-06 (topic provenance)