- ChatGPT will now use The Washington Post to provide answers using the newspaper’s reporting
- ChatGPT answers will now include summaries, quotes, and article links
- The move builds on OpenAI’s growing roster of news partners and The Post’s own AI initiatives
The Washington Post has inked a deal with OpenAI to make its journalism available directly inside ChatGPT. That means, the next time you ask ChatGPT something like “What’s going on with the Supreme Court this week?” or “How is the housing market today?” you might get an answer including a Post article summary, a relevant quote, and a clickable link to the full article.
For the companies, the pairing makes plenty of sense. Award-winning journalism, plus an AI tool used by more than 500 million people a week, has obvious appeal. An information pipeline that lives somewhere between a search engine, a news app, and a research assistant entices fans of either or both products. And the two companies insist their goal is to make factual, high-quality reporting more accessible in the age of conversational AI.
This partnership will shift ChatGPT’s answers to news-related queries so that relevant coverage from The Post will be a likely addition, complete with attribution and context. So when something major happens in Congress, or a new international conflict breaks out, users will be routed to The Post’s trusted reporting. In an ideal world, that would cut down on the speculation, paraphrased half-truths, and straight-up misinformation that sneaks into AI-generated responses.
This isn’t OpenAI’s first media rodeo. The company has already partnered with over 20 news publishers, including The Associated Press, Le Monde, The Guardian, and Axel Springer. These partnerships all have a similar shape: OpenAI licenses content so its models can generate responses that include accurate summaries and link back to source journalism, while also sharing some revenue with publishers.
For OpenAI, partnering with news organizations is more than just PR polish. It’s a practical step toward ensuring that the future of AI doesn’t just echo back what Reddit and Wikipedia had to say in 2021. Instead, it actively integrates ongoing, up-to-date journalism into how it responds to real-world questions.
WaPo AI
The Washington Post has its own ambitions around AI. The company has already tested ideas like its “Ask The Post AI” chatbot for answering questions using the newspaper’s content. There’s also the Climate Answers chatbot, which the publication released specifically to answer questions and share information based on the newspaper’s climate journalism. Internally, the newsroom has been building tools like Haystacker, which helps journalists sift through massive datasets and find story leads buried in the numbers.
Starry-eyed idealism is nice, but there are some open questions. For instance, will the journalists who worked hard to report and write these stories be compensated for having their work embedded in ChatGPT? Sure, there’s a link to their story, but that doesn’t count as a view or help lead a reader to other pieces by the author or their colleagues.
From a broader perspective, won’t making ChatGPT a layer between the reader and the newspaper simply continue the undermining of subscription and revenue necessary to keep a media company afloat? Whether this is a mutually supportive arrangement or just AI absorbing the best of a lot of people’s hard work while discarding the actual people remains to be seen. Making ChatGPT more reliable with The Washington Post is a good idea, but we’ll have to check future headlines to see if the AI benefits the newspaper.
Leave a comment