Meta Changed Its Mind

You've probably heard the line: "Facebook killed the news." Starting around 2021, Meta made a deliberate push to deprioritize news on its platforms. It shut down the Facebook News tab, tweaked its algorithm to suppress publisher posts in the feed, and sent a clear signal to the industry: we're done with journalism. Traffic dried up. The consensus in media circles was that Meta had simply moved on.

Then, in December 2025, Meta did a complete 180. The company announced multi-year licensing deals with seven major news organizations — CNN, Fox News, USA Today, and others — agreeing to pay for the right to use their content. The exact figures weren't disclosed, but industry observers described the deals as "substantial."

What changed? The answer is simple: AI chatbots.

What's Actually in These Deals

The roster of publishers Meta signed is worth a look:

• CNN — global breaking news
• Fox News, Fox Sports — conservative news and sports
• USA Today, USA Today Network — national and regional U.S. news
• Le Monde Group — France's leading newspaper
• People Inc. — entertainment and lifestyle
• The Daily Caller, The Washington Examiner — conservative media

The political and editorial range is deliberately broad. The deals are described as "multi-year," but the duration and dollar amounts remain confidential.

Here's how it works: when a user asks Meta AI a news-related question, the chatbot pulls from these publishers' articles and surfaces an answer — with a source link at the bottom so users can read the original piece. The feature rolls out across Facebook, Instagram, WhatsApp, and Messenger — essentially Meta's entire product surface — in more than 200 countries. That means potentially billions of people could start getting news through Meta AI.

Why AI Chatbots Need News in the First Place

If you've used ChatGPT, you know the frustration: ask it about something that happened last week and it draws a blank. Its training data has a cutoff, so it simply doesn't know. That's why OpenAI struck licensing deals with the Financial Times, Associated Press, and Axel Springer — and gives paying subscribers access to real-time web search.

Google's Gemini has a built-in advantage: it's backed by the world's largest search engine, so Google News is already within reach. No separate deals needed.

Meta has no search engine. That means the only way to give Meta AI access to breaking news is to go directly to the source — sign the deals, pay for the content, and build the pipeline. Otherwise, Meta AI just can't answer questions about anything that happened in the last few months.

Here's how the major AI chatbots compare on news strategy:

Perplexity's approach — scraping publisher content without permission, summarizing it, and presenting the results as its own — landed it in legal trouble with the New York Times and Chicago Tribune, among others. The argument: you're taking our reporting, repackaging it, and capturing the traffic that should be ours.

The industry is clearly moving toward formal licensing. Meta is following that playbook.

For Publishers: Opportunity or Trap?

Whether this is good news for publishers depends on how you frame it.

The upside:

• Licensing revenue — real money from a platform that reaches billions of users.
• Brand visibility — every time Meta AI cites "according to CNN," that's a brand impression.
• A new traffic funnel — users who click the source link end up on the publisher's site.

The downside:

• If the AI summary is good enough, nobody clicks. Why read the full article when the chatbot already told you what happened?
• Less traffic means lower ad revenue — which, for most publishers, still dwarfs licensing income.
• Long-term platform dependency: your content becomes a feed for Meta's product, with Meta controlling how it's surfaced and summarized.

The developer community has been blunt about this: "Source links don't matter if nobody clicks them. If the AI summary is good enough, there's no reason to go to the original."

Research has shown that ChatGPT cites Wikipedia and Reddit more than almost any other source — yet neither Wikipedia nor Reddit has reported a meaningful traffic bump as a result. Users read the AI's answer and move on.

What This Means for Journalism

In the near term, expect more publishers to chase AI licensing deals. Microsoft's February 2026 launch of its "Publisher Content Marketplace" — a platform where news organizations can sell their content directly to AI companies — is a sign of where this is heading. Both sides get something: AI companies get clean, legally licensed data; publishers get a new revenue stream.

In the medium term, if AI summaries keep getting better, fewer people will bother reading full articles. Ask ChatGPT to summarize a news story today and you'll get a tight three-sentence recap in seconds. Why scroll through a 1,500-word piece if the bot already nailed the key points?

In the long term, there's a real risk that journalism slides into becoming an AI training data provider — reporters doing the work of reporting, and AI capturing the value of distribution. That hollows out the economic model that makes investigative and accountability journalism possible.

The irony is hard to miss: AI needs news to function, while simultaneously disrupting the business model that produces news.

There's a counter-argument, though. In a world drowning in AI-generated slop and low-quality content, verified, sourced reporting becomes more valuable, not less. "This came from CNN" or "the NYT investigated this" carries real signal in an environment where provenance is everything.

Maybe AI isn't killing journalism. Maybe it's forcing journalism to evolve.

What You Should Take Away From This

Whether you're using Meta AI or ChatGPT, get into the habit of checking the source. "The AI said so" is not a standard of evidence. "According to [source]" should be your first prompt — not your last.

And occasionally, click through. AI summaries strip out context, analysis, and the interview quote that changes the whole picture. AI extracts facts; journalism explains why they matter.

Meta's move is just the beginning. Google, OpenAI, and every other major AI player will go down a similar path. News has become data and product in the AI economy. Publishers will have to adapt — or gradually become irrelevant.

News Isn't Dying. It's Transforming.

Meta isn't paying for news out of goodwill. It needs verified, real-time content to compete in the AI chatbot race — full stop. ChatGPT, Gemini, Meta AI: they're all fighting over the same thing. Current, credible information. That's the new battleground.

Publishers get licensing revenue; they also inherit the risk of becoming permanently dependent on AI platforms to reach their audience. Readers get convenience; they also risk developing a shallow news diet built entirely on AI-curated summaries.

Next time you're reading a Meta AI or ChatGPT response about something in the news, click the source link. It's a small act that keeps the journalism ecosystem alive.

The form of journalism will change. The value of verified information won't. In a world full of AI-generated content, "a real journalist reported this" may end up being the most powerful differentiator of all.

Sources

- Meta official announcement
- TechCrunch: Meta signs commercial AI data agreements
- Axios: Meta strikes multiple AI deals
- CNBC: Meta strikes AI deals with publishers
- Digiday: Meta enters AI licensing fray