OpenAI's Publisher API: How News Sites Let the Wolf Into the Henhouse

• OpenAI launched the Publisher API in June 2024, targeting newsrooms directly.
• Major publishers like The Associated Press and Axel Springer have already inked deals.
• API-accessed content is directly ingested and summarized by ChatGPT, bypassing source sites.

Most news executives still believe OpenAI’s Publisher API is “a new traffic source.” I’ll spell it out: when you let OpenAI ingest your content directly, you’re serving your own audience to the beast—and then acting surprised when your pageviews crater. The Associated Press, Financial Times, and Axel Springer sold their archives for licensing checks so small they wouldn’t cover a new CMS migration. This isn’t the “partnership” PR flacks are hyping; it’s the same Faustian bargain the music industry made with Spotify, except newsrooms will get even less out of it.

Let’s be real about what the Publisher API actually does: it gives ChatGPT the power to summarize, repackage, and answer queries using your content, with only a “source” link as your prize. No user is clicking that. Remember when Google’s featured snippets nuked search traffic for recipe blogs and medical publishers? Now, OpenAI is doing it at scale, and the average news CMS is so slow and ad-riddled it’s handing Sam Altman every excuse to keep users inside ChatGPT. The API gives OpenAI a clean, structured firehose; in return, you get… a mention, if you’re lucky, and a tiny payout. That’s the deal lazy publishers are rushing to sign.

The worst offenders are the agencies and SEO “consultants” cheering this on. You’ll find the same LinkedIn gurus who told you to chase “keyword density” in 2022, now peddling Publisher API integration as if it’s a gold rush. They’re either too lazy or too ignorant to realize that every API call is another nail in their client’s traffic coffin. Here’s the metric that matters: after Google’s AI Overviews, publishers saw an average 18% drop in organic search traffic. With OpenAI’s direct ingestion, expect that trend to double—except now you can’t even optimize your snippet.

There is one uncomfortable solution, but nobody wants to hear it: block OpenAI’s crawler and refuse their API. Yes, you’ll lose whatever pennies these “partnerships” promise, but you might keep your direct audience and brand intact. If every publisher had the spine to shut the door, OpenAI would have to pay—actually pay—for content. But most newsrooms, driven by lazy agency logic and quarterly desperation, will let themselves be swallowed whole, and call it innovation.

Frequently Asked Questions

What is OpenAI’s Publisher API and how does it work?

OpenAI’s Publisher API lets news organizations send their articles and metadata directly to ChatGPT, which can then summarize and answer user questions using this data. In exchange, publishers get source attribution and a licensing fee, but lose control over how their content is presented and monetized.

Why is this API a threat to publisher traffic and revenue?

When ChatGPT uses your content to answer user questions directly, there’s little incentive for users to click through to your website. This leads to plummeting pageviews and reduced ad revenue—just like what happened with Google’s featured snippets, only now it’s powered by LLMs with even fewer attribution clicks.

What should publishers do to protect their interests?

Publishers should block OpenAI’s crawler and refuse to participate in the Publisher API until there are clear, enforceable terms guaranteeing meaningful compensation and control. Otherwise, they risk becoming just another invisible data source in the AI supply chain—and losing their audience for good.