What happened: News Corp has struck a multiyear AI content-licensing agreement with Meta, letting the Facebook and Instagram owner use parts of News Corp’s US and UK journalism to help train its AI products.
Why it matters: The deal is another data-point in the shift from publishers being scraped for free to publishers trying to price access — treating news as a paid input into model training rather than an output to be summarised away by AI systems.
Wider context: Meta is spending heavily on AI infrastructure, while media groups are split between licensing, partnership, and litigation strategies as AI features in search and social feeds threaten click-through traffic and ad revenue.
Background: News Corp previously signed a major agreement with OpenAI in 2024, and its CEO has framed the company’s stance as “woo or sue” — collaborate when there’s a deal on the table, and pursue legal action when content is taken without permission.
News Corp is essentially an AI ‘input company’, chief executive says, after US$150m deal with Meta — The Guardian
Singularity Soup Take: Calling journalism an “input” is honest — but the real fight is over who captures the value when AI features reduce referral traffic; licensing cheques help, yet they may also normalise a future where publishers get paid while audiences quietly disappear.
Key Takeaways:
- Scope Matters: The reported Meta licence covers News Corp content in the US and UK (including outlets such as the Wall Street Journal and New York Post), while Australian mastheads are explicitly excluded from the arrangement.
- Pricing The Feedstock: The Guardian reports the Meta deal is worth up to US$50m a year and is expected to run at least three years, signalling that training rights are being negotiated like other long-term supply contracts.
- Carrot And Stick: News Corp’s CEO described a “woo or sue” approach to AI firms — pursue partnerships when terms are acceptable, but reserve litigation for unlicensed use — as publishers test what leverage they actually have.