SAN FRANCISCO –
News Corporation has struck an artificial-intelligence content-licensing agreement with Meta valued at up to US$50 million a year that will allow the social media company to use News Corp’s US and UK content to train its AI products. The deal, expected to run for at least three years, excludes News Corp’s Australian mastheads.
The agreement formalises a recurring revenue stream for the publisher group while expanding the supply of professionally produced news and data available to large AI models. Company executives say the arrangement sits alongside an existing five-year, US$250 million licensing deal News Corp signed with OpenAI in 2024.
The licensing arrangement transfers a defined commercial entitlement to use News Corp content into AI training pipelines and alters the terms under which platform operators access premium news material. For News Corp, the transaction creates an additional income line distinct from subscriptions and advertising; for Meta, it secures a curated source of breaking news, commentary and proprietary reporting to feed model development and product features across Facebook and Instagram.
Deal terms and scope
News Corp’s portfolio titles that are within the scope of the agreement include Dow Jones and several major US and UK newspapers; several Australian mastheads are explicitly excluded from the deal. The reported headline value is up to US$50 million per year, with a minimum expected duration of three years. People familiar with the talks said the agreement covers both archival and ongoing output, giving Meta access to a rolling feed of professionally edited journalism rather than a static data dump.
| Counterpart | Headline value | Duration | Territorial scope |
|---|---|---|---|
| News Corp | Up to US$50m per year | Expected at least three years | United States and United Kingdom content |
| News Corp – earlier agreement | US$250m | Five years (signed 2024) | Content licensed to OpenAI |
| Meta – infrastructure investment | Up to US$6bn | Contract announced January 28, 2026 | AI datacentre fibre-optic supply |
While financial details have not been fully disclosed, analysts say the headline value and multi-year tenor are likely to be closely watched by other publishers seeking benchmarks for the price of training data in commercial AI models.
Corporate positioning and strategy
Robert Thomson, News Corp’s global chief executive, framed the company’s role in AI product development as that of a supplier of high-quality inputs for model training. He characterised the publisher’s content as a corporate asset that can be monetised through licensing arrangements with technology platforms while retaining legal remedies where use is unauthorised.
“We’re essentially an input company. The great threat in the age of AI is going to be to what you might call output companies. We’re an input in the way that semiconductors are an input, in the way that datacentres are an input, in the way that energy is an input. You look at breaking news, you look at unique real estate information.” – Robert Thomson
Thomson further said he favours a “woo or sue” approach toward AI firms – pursuing commercial deals where possible and litigation where content is used without permission – and noted ongoing direct lines of communication with senior executives at major AI and social platforms, stating that “Mark and I converse on a pretty regular basis, across WhatsApp, obviously”.
News Corp’s holdings include Dow Jones, the New York Post and other titles in the US and UK that provide time-sensitive reporting and proprietary data. The corporate strategy evident in the new deal seeks to convert that informational capital into contracted revenue from technology firms that are investing heavily in model development and AI infrastructure. For Meta, the arrangement is intended to underpin generative features that summarise, recommend or contextualise news within its apps, while offering an answer to criticism that AI products can surface unverified or low-quality material.
Market and regulatory context
The transaction arrives as major technology platforms expand both compute capacity and supply-chain investments for AI. Meta announced a multibillion-dollar infrastructure arrangement with Corning on January 28, 2026, to supply fibre-optic cable for datacentres, illustrating the scale of upstream spending behind AI product ecosystems.
At the same time, major publishers have taken divergent approaches to the use of their content by AI companies. News Corp’s licensing deals sit alongside legal action from other outlets against AI developers and cloud providers over content use. Publishers and platforms operate within overlapping regulatory regimes for copyright, data protection and competition in the US, the UK and the EU; commercial agreements such as this one alter the economic incentives around content distribution and model training.
In the US and Europe, regulators are sharpening their focus on how training datasets are assembled and whether rights holders are fairly compensated. Draft and enacted AI rules, including the emerging regime under the EU Artificial Intelligence Act, are beginning to codify obligations around transparency, data provenance and risk controls for high-impact AI systems. For large publishers and platforms, voluntary licensing deals are increasingly being assessed not only as revenue generators but also as tools to demonstrate compliance and mitigate litigation or enforcement risk.
Precedents and corporate governance implications
The licensing approach mirrors a broader pivot among legacy news organisations toward monetising content directly with AI firms rather than relying solely on ad-supported distribution via platforms. News Corp’s deal structure and public statements emphasise controlled commercialisation and legal recourse as governance mechanisms to protect editorial assets.
The agreement also reinforces an industry trend in which platform firms secure licensed content sets to improve product quality while publishers seek diversified revenue that is less dependent on referral traffic. News Corp’s existing, separately negotiated arrangement with an AI developer and other strategic partnerships across the sector provide precedents for contractual terms, revenue recognition and content-use limitations that corporate auditors and boards will assess under accounting and disclosure standards. Company directors are likely to scrutinise whether revenue-sharing formulas, usage caps and audit rights are robust enough to capture upside if AI products scale rapidly.
News Corp has contrasted its licensing and partnership strategy in international markets with a more adversarial stance taken by parts of its Australian business. Separately, other media groups have pursued legal challenges or strategic partnerships with AI developers at different commercial and legal risk tolerances; one major UK media group announced a strategic partnership with an AI developer on February 14, 2025.
Executives and boards will now weigh the revenue benefits of multi-year licensing agreements against potential impacts on subscription flows, direct audience relationships and regulatory scrutiny. The licensing model also channels new counterparty exposure onto publishers’ balance sheets and contracts teams, requiring governance attention to indemnities, content-use definitions and audit rights. For public companies, those questions feed into disclosure decisions and risk reporting, including how to describe AI-related dependencies and intellectual-property strategies in annual filings to the US Securities and Exchange Commission and other market regulators.
The agreement is expected to run for at least three years.
Keep reading
