In the last few months, news organizations have leapt into bed with OpenAI, hatching Faustian bargains where the cash-strapped media industry exchanges a monetary pittance for OpenAI’s right to scrape and integrate their content into things like ChatGPT. Those that have signed in blood include News Corp (publisher of the Wall Street Journal), the Financial Times, People magazine publisher Dotdash Meredith, the AP, and now, as Axios reports, The Atlantic and Vox Media.
The Atlantic quickly confirmed this new deal in a statement on its site. The publication says that it’ll be a “premium news source” in OpenAI and that all its citations will be clearly attributed to The Atlantic with links back to the original content. There are concerns from publishers that users of AI chatbots don’t actually need to go to the original sources; perhaps the calculus is that, for an industry in the twilight of its lifespan, some inbound link traffic is better than none. Then again, by agreeing to be scraped at all, perhaps The Atlantic is effectively wading directly into the tarpit of its own extinction (and of media as a whole).
There will also be an experimental “microsite” called Atlantic Labs that’ll showcase “new products and features to better serve its journalism and readers.” There’s no indication yet that that content will involve anything created directly by AI, as sites like CNET and Sports Illustrated have tried with disastrous results.
According to Axios, Vox Media (publisher of its flagship news site Vox, tech site The Verge, the network of sports blogs under the SB Nation banner and many more) will have a similar style of attribution for its content. We’ve reached out to the company and will update this story if we get any confirmation or details on its deal with Open AI.
Vox Media is also said to be using OpenAI data both internally and in public-facing content. Axios mentions using OpenAI to “match shoppers” with content endorsed by Vox’s many commerce channels, like The Strategist section of New York Magazine. Internally, Vox will use OpenAI to as part of its advertising platform to tweak ad creative to make visitors more likely to click; it’ll also help them target ads at people likely to buy stuff from them. Details on how both of these programs will work is minimal right now, though.
While a number of publishers have been quick to embrace AI, not everyone is so enthused. The New York Times sued both OpenAI and Microsoft for copyright infringement in December, saying that both companies use its material without permission to train their models. More recently, eight publications owned by the Alden Capital Group, including the Chicago Tribune and New York Daily News, sued both companies with a similar complaint. At this point, it seems like it’s either spend time and money in a lawsuit to go after OpenAI’s rampant intellectual theft or cut a deal that’ll make you some spending cash in a dire media market.
It was only last week The Atlantic published its own screed decrying media organizations which had taken petty cash from AI interlopers in exchange for something of significantly greater value. The odds unfortunately suggest this story (and my moral high ground) will age just as poorly in the near future.