Seattle Times sues OpenAI. Get AI indemnification in writing.
Seattle Times and Newsday sued OpenAI and Microsoft over training data — after taking their funding. Your AI vendor relationship needs contract terms, not goodwill.
The Seattle Times and Newsday filed a federal copyright and trademark complaint against OpenAI and Microsoft in the Southern District of New York, calling AI products "rapacious consumers" of the journalism they were trained on. The detail worth your attention is buried in TechCrunch's coverage: Microsoft and OpenAI had previously funded some of the Seattle Times' journalism projects and fellowships. The partnership did not prevent the lawsuit. If you are building on a frontier model, your AI indemnification terms matter more than your vendor relationship.
What actually happened
The two newspapers allege OpenAI and Microsoft scraped their sites — including content behind paywalls — and used the articles to train and operate ChatGPT, Copilot, and Bing's AI features. The complaint's framing is blunt: "AI products like ChatGPT and CoPilot are touted as producers of content, but in fact they are rapacious consumers." Engadget reports the publishers are asking the court to order the destruction of copies of their works along with the training datasets and models that incorporate them.
Microsoft's response, per TechCrunch, was that it is "surprised by the lawsuit" but "always happy to sit down and explore solutions to this type of dispute." That is the sound of a business relationship that did not survive contact with a legal department.
This is the same defendant pair the New York Times sued in December 2023, and it lands in the same month Sony Music Publishing and Warner Chappell went after Anthropic and its founders personally. The pattern is no longer a pattern. It is the operating environment.
Why it matters for your business
You are not going to be sued for using ChatGPT. But you can absolutely be the downstream party when a court orders a model retired, a dataset destroyed, or an output pulled. And you can be the one holding the bag if a customer claims your AI-generated product copy, support macro, or listing description infringes something.
So read the paper. Most enterprise AI terms — OpenAI's Copyright Shield, Microsoft's Customer Copyright Commitment, Google's generative AI indemnity — offer output indemnification with conditions attached: you must be on a paid tier, you must not have disabled the safety filters, you must not have deliberately prompted for infringing content. Free tiers and consumer plans usually get nothing. If you are running production workloads through a personal API key on a hobby plan, you have no coverage and no counterparty.
Three moves, none of them expensive. One: pull the actual indemnification clause for every AI vendor you send customer data or ship output from, and put the conditions in a one-page file. Two: keep the model layer swappable, so a legal shock at one vendor is a config change and not a rebuild. Three: log which model version produced what, so if an output is ever challenged you can answer with a record instead of a shrug. Goodwill funded the Seattle Times' fellowships. It did not stop the filing.
Key takeaways
- Seattle Times and Newsday sued OpenAI and Microsoft in the Southern District of New York over training on their journalism, including paywalled content
- Microsoft and OpenAI had previously funded some of the Seattle Times' journalism projects — a commercial relationship did not prevent litigation
- The publishers are asking for destruction of the infringing datasets and models, per Engadget
- Vendor output indemnification is conditional: paid tiers, filters left on, no deliberate infringing prompts — read your actual clause
- Keep the model layer swappable and log which model version produced which output
Running AI in production without reading the terms? We build vendor-agnostic systems with a swappable model layer and an output audit trail, so a lawsuit at your vendor is a config change instead of a rebuild. See how we build systems you own or bring us your stack.
Sources: TechCrunch, Engadget.
- #ai-vendors
- #copyright
- #indemnification
- #contracts
- #risk
Tommy Rush — Founder, Rush Commerce
Operator turned builder. 15+ years running operations — now shipping the systems businesses run on. More
Get The Rush Report weekly — one email, zero fluff.
Keep reading
Stop Rogue AI Act wants a machine-readable agent inventory
The Stop Rogue AI Act directs NIST to set AI agent security standards — machine-readable inventory, tamper-proof action logs, and teeth for federal contractors.
Read itSaudi Arabia's humain-m3 runs on Chinese open weights
HUMAIN's 428B Arabic model humain-m3 is built on MiniMax-M3 and will ship under the MiniMax Community License. Read the license before you design around it.
Read it