TRV-2026-0703Version 1 · Certified
Reason for this version
Certified into the record
Canonical text (the exact bytes fingerprinted)
TRUVACE RECORD VERSION record: TRV-2026-0703 version: 1 kind: certified reason: Certified into the record timestamp: 2026-08-09T06:32:24.398294Z status: published lens: g_space sector: entertainment headline: News Corp is essentially an AI ‘input company’, chief executive says, after US$150m deal with Meta dek: News Corp’s global chief executive has described news organisations as a valuable “input” for artificial intelligence, as the media empire signs an AI content licensing deal with Meta worth up to US$50m (A$71m) a year. In an upbeat presentation, the chief executive of Rupert Murdoch’s company, Robert Thomson, said the “reliable” breaking news and information in publications like the Australian, the Times of London and Dow Jones was “hard to beat” as an “input” for AI. The Meta deal, which was revealed by the Mur… gain_title: News Corp licensed its US and UK news content to Meta for AI training in a deal worth up to $50m per year for at least three years, positioning its journalism as a paid input for AI systems. problem_title: (none) trace_subject: (none) gain_reading: News Corp licensed its US and UK news content to Meta for AI training in a deal worth up to $50m per year for at least three years, positioning its journalism as a paid input for AI systems. gain_evidence: worth up to US$50m (A$71m) a year | allow Facebook and Instagram's parent company to scrape News Corp's US and UK content to train its artificial-intelligence products | expected to last at least three years problem_reading: (none) problem_evidence: (none) quick_read: On 4 March 2026, The Guardian reported that News Corp chief executive Robert Thomson called news organisations an AI input company while announcing a licensing deal with Meta worth up to US$50m a year for at least three years to let Meta scrape US and UK content to train AI products. The deal matters because it shows a major publisher monetizing its archive as AI training data while the wider news industry worries that AI search integration cuts referral traffic; it remains uncertain whether licensing revenue offsets long-term audience loss and how far the model extends given Australian titles were excluded and staff concerns about internal AI tools. limitation: tag: Evidence-backed gain key_points: News Corp chief executive Robert Thomson described news organisations as an "input company" comparable to semiconductors, datacentres and energy. | The Meta deal covers outlets including the Wall Street Journal and New York Post, but Australian mastheads like the Daily Telegraph and Herald Sun are excluded. | Thomson previously signed a $US250m five-year deal with OpenAI in 2024 and said he takes a "woo or sue" approach to AI companies. | News Corp Australia has introduced an in-house AI tool called "NewsGPT" that some journalists have expressed concern about. rundown: Thomson made the remarks at a Morgan Stanley tech conference in San Francisco on Monday, saying reliable breaking news and real estate information from titles like the Australian, the Times of London and Dow Jones was hard to beat as an AI input. The context includes Meta's multibillion-dollar AI infrastructure push, including a deal worth up to $US6bn with Corning for fibre optic cables for datacentres, and contrasting industry responses such as the New York Times suing OpenAI and Microsoft over training use. sources: - journalism | The Guardian | https://www.theguardian.com/media/2026/mar/04/news-corp-meta-ai-deal-us50m | 2026-03-04 prev: 0000000000000000000000000000000000000000000000000000000000000000
- sha256
- fd40e6f0beeb7b32c4ffd850eaf3f01377fe0bb4540456662baa1bb933ebe71b
- previous
- 0000000000000000000000000000000000000000000000000000000000000000
Verify this record
How to verify without trusting this page
Fetch the canonical text of any version from /api/record/TRV-2026-0703 and hash it yourself — for example shasum -a 256 on the saved canonical field. The result must equal content_hash, and each version’s text ends with prev:followed by the prior version’s hash (version 1 chains to 64 zeros). If a single character of any version had been altered since certification, the chain would not reproduce.
ace