Tomorrow's Publisher — forecast register and claim audit 28 October 2026 Article: Future shows why publishers need more than a place in an AI answer Author: Ivan Massow The article is an editorial judgement supported by checked sources and Noah's automatically supplied quick forecasts. Probabilities are estimates of the exact outcomes below, not observed rates of business success. All five final estimates are retained. Four appear in the article; the fifth is excluded for the specific baseline/predicate issue described below. FORECASTS — Origin: Noah final quick forecast publisher_chatgpt_metrics_20270331 Question: Will any of Future plc, Financial Times Group, Guardian Media Group or News Corp publish, by 2027-03-31, a numeric measure of ChatGPT-originated referrals, engagement or revenue for one of its UK publisher properties? YES settlement: By 2027-03-31, at least one named publisher has published a numeric ChatGPT-originated referral, engagement or revenue measure for a UK property in a company report, results release, executive presentation or other attributable public statement. NO settlement: By 2027-03-31, none of the named publishers has published such a numeric ChatGPT-originated measure for a UK property. Final YES: 41%. Final NO: 59%. YES uncertainty interval: 22–62%. Deadline: 2027-03-31. Native run reference: 14f33d845fd4c0bfaed1eec966bc903e531865510547ce7ad6bc289af1118eab. Final receipt: 20981c1988b1364310bd9f7d015329e6c45eb553d1b7a64e3b17621b114ea81f. Final calculation completed: 2026-10-28T05:31:58.570622+00:00. Placement: Article. Scope: this is publication of a numeric ChatGPT-originated measure for a UK publisher property by one of four named organisations, not evidence of profitability or a forecast that no private benefit occurs. The final explanation interprets the eligible publication window as after the 28 October 2026 forecast origin; that interpretation is preserved here alongside the original unchanged contract. publisher_chatgpt_metrics_20271028 Question: Will any of Future plc, Financial Times Group, Guardian Media Group or News Corp publish, by 2027-10-28, a numeric measure of ChatGPT-originated referrals, engagement or revenue for one of its UK publisher properties? YES settlement: By 2027-10-28, at least one named publisher has published a numeric ChatGPT-originated referral, engagement or revenue measure for a UK property in a company report, results release, executive presentation or other attributable public statement. NO settlement: By 2027-10-28, none of the named publishers has published such a numeric ChatGPT-originated measure for a UK property. Final YES: 59%. Final NO: 41%. YES uncertainty interval: 40–74%. Deadline: 2027-10-28. Native run reference: a7e82487a2b1a3521b78abcd2c743ac6b147f3a6f63e66f387c058745e55a764. Final receipt: 0e1905571d1f1014391dcefcc530c713f25abe97cdcd12977bd1a9c9cc82c789. Final calculation completed: 2026-10-28T05:29:39.757483+00:00. Placement: Article. Scope: this is publication of a numeric ChatGPT-originated measure for a UK publisher property by one of four named organisations, not evidence of profitability or a forecast that no private benefit occurs. The final explanation interprets the eligible publication window as after the 28 October 2026 forecast origin; that interpretation is preserved here alongside the original unchanged contract. openai_publisher_analytics Question: Will OpenAI make available to its contracted news-publisher partners, by 2027-10-28, an analytics product that separately reports both appearances or citations in ChatGPT answers and outbound clicks to publisher domains? YES settlement: By 2027-10-28, OpenAI has made such a product available to contracted news-publisher partners, with separate reporting for answer appearances or citations and outbound publisher clicks, as evidenced by OpenAI product documentation or attributable partner demonstrations or statements. NO settlement: By 2027-10-28, OpenAI has not made a publisher analytics product meeting both separate-reporting requirements available to contracted news-publisher partners. Final YES: 43%. Final NO: 57%. YES uncertainty interval: 25–62%. Deadline: 2027-10-28. Native run reference: 69d557cbe29ad302697a51da22fd0705c9d28fb086ee176c58e8707a0996e0f5. Final receipt: 18becb42d458b636cc13e38e91dc5775aca8f7aa978a066e0b3651e082eaf7e2. Final calculation completed: 2026-10-28T05:29:52.260945+00:00. Placement: Article. Scope: partner access to a product reporting both answer appearances/citations and outbound clicks. Ordinary tagged referrals, advertiser analytics or a product announcement without availability are insufficient. openai_deal_break Question: Will Financial Times Group or News Corp terminate or decline to renew a publisher-content licensing or supply agreement with OpenAI by 2027-10-28? YES settlement: By 2027-10-28, Financial Times Group or News Corp has terminated, or allowed to expire without renewal, a publisher-content licensing or supply agreement with OpenAI, as established by the companies' releases, filings or attributable statements. NO settlement: By 2027-10-28, neither Financial Times Group nor News Corp has terminated or declined to renew such an agreement with OpenAI. Final YES: 16.2%. Final NO: 83.8%. YES uncertainty interval: 7–30%. Deadline: 2027-10-28. Native run reference: 13568cb057fba23f616860254dc421900245be3a6012e2a577a446e9312c635b. Final receipt: 8cc0e72d38b0f24a67b4808964288a67a90a6f16ddddcc96effd365be2c181df. Final calculation completed: 2026-10-28T05:30:18.351469+00:00. Placement: Article uses the exact binary NO complement, 83.8%, with NO interval 70–93%. This means neither of the two named groups has a qualifying break; it does not mean all publishers remain, no renegotiation occurs, or the terms are unchanged. ai_publisher_remedy Question: Will the European Commission or UK Competition and Markets Authority issue, by 2027-10-28, a final legally operative rule, order or decision requiring a major generative-AI provider to obtain publisher consent or remunerate publishers for using news content in generated answers? YES settlement: By 2027-10-28, the European Commission or UK Competition and Markets Authority has issued a final legally operative rule, order or decision imposing a publisher-consent or remuneration requirement on OpenAI, Google, Microsoft, Anthropic, Perplexity or another major generative-AI provider for news content used in generated answers. NO settlement: By 2027-10-28, neither authority has issued a final legally operative measure meeting those requirements; investigations, consultations, preliminary findings and voluntary agreements do not qualify. Final YES: 55%. Final NO: 45%. YES uncertainty interval: 27–99%. Deadline: 2027-10-28. Native run reference: ecad429e5ec92ccd78cc51b8b2f24adfdce05e3bad1821ed9a9188843c3e4e64. Final receipt: 0333a12bb931a096b6d8ea9e7cc061eec59443a3d330dd2ad8a8d5a3472c157b. Final calculation completed: 2026-10-28T05:30:07.942329+00:00. Placement: Audit only. The CMA imposed a binding publisher-control requirement on Google on 3 June 2026. Whether opt-out controls satisfy this target's publisher-consent wording is unresolved; an affirmative opt-in/payment requirement is a narrower reading. The final calculation acknowledges both interpretations. This is not printed as a clean forecast of a new future rule. Its December 2026 implementation reference is not adopted as an observed event as of today. TIMING The two disclosure forecasts retain the same named-publisher, UK-property and metric conditions at the two tested deadlines. YES is 41% by 31 March 2027 and 59% by 28 October 2027. These are separate dated estimates, not a fitted crossing date; no untested intervening odds are inferred. Their intervals are 22–62% and 40–74%, respectively. SOURCE AND CLAIM CHECKS 1. David Konitzny's primary LinkedIn post describes 12 observed providers and 43 domains, not a complete provider census. Its 1,000-list-summary sample concerns inclusion of list entries in pre-generated summaries, not final recommendation probabilities, clicks or causal partner advantage. The Decoder dates its coverage to 27 October. The original post is a technical observation, not an OpenAI announcement. 2. Future's OpenAI deal was announced 4 December 2024. The 200-plus-brand description belongs to that historical announcement, not a verified current portfolio count. 3. Future's 14 May 2026 release concerns the half year ended 31 March. Its 8% direct-ad growth, direct ads twice programmatic and 16% of total revenue directly correlated with website sessions are company-reported measures. They neither isolate the OpenAI agreement nor establish immunity from traffic losses. 4. Tom's Guide's Product Finder, expert Q&As, newsletters and 78,000 club members are from Future's 15 April 2026 relaunch announcement. No subsequent measured conversion uplift is asserted. 5. PLS's October update distinguishes AI product licensing (350-plus publishers opted in; no significant short-term revenues expected) from workplace AI licensing (800-plus participating publishers; payments since March). Counts may overlap; they must not be summed or relabelled ChatGPT partners. Licensing options, actual licences and actual payments remain different stages. 6. OpenAI's current publisher FAQ describes ordinary referral tracking with utm_source=chatgpt.com. It also says any public website can appear in ChatGPT search. A publisher does not have to hold a content deal to be discoverable. Search discovery and model-training controls differ. 7. The FT and Guardian partnership announcements establish attributed summaries/links and intended product or audience benefits, not a quantified return. 8. The CMA's official publisher-conduct page records imposition on 3 June 2026 and includes generative-AI use controls, engagement metrics, attribution and access. Google's duties are not automatically OpenAI's duties. Nothing in the article promises a general remuneration floor. 9. The laptop example is explicitly hypothetical. Editorial independence and measurement recommendations are editorial judgements, not claims of a measured platform-wide causal effect. RESEARCH METHOD Initial programme: nabdl_programme_run_58e415cc168346c639698d2b3698645b49a0c43e8b924a11 Required jobs: nabdl_job_5f3d79364af70fedcf900f2a7b104968ba4627e10a65dea543fda9af3d83fdae, nabdl_job_d8f74aab05a7110c5f583f99046ad27546c8c7745102dd89d41fb3cef6b3b150 The initial programme completed 33 of 35 planned checks across 28 distinct lines; two unfinished checks remain outside conclusions. Seven of 12 completed hypothesis tests admitted no directly relevant evidence. That is a bounded evidence gap, not falsification or proof of absence. The lexical-match trajectories and retained-card counts are not adoption, traffic, revenue or incident rates and were not printed as such. Off-topic metrology, infrastructure and insurance-M&A records were excluded. Music licensing, PR placement and generic distribution stories do not establish a news-publisher ChatGPT return. All five automatic native targets reached completed final reassessment after a number-blind host contribution with checked public sources. No standalone forecast job, extra forecast lane, revised deadline or host replacement probability was used for the article. Native receipt digests and final computation receipts remain in the private archive. Post-draft investigation: nabdl_job_d39bff7a679dac5f0eb8fda480c07f9e8d2f4bf9844cc06c36c0bec3a10704d6 completed all six paired checks. No matched checkout-commission outcome, opt-in-to-payment cohort or same-URL correction-propagation trace was established. Four census angles failed and are not treated as measured silence. Contextual marketplace material led to Microsoft's 3 February primary announcement: publisher-defined terms and usage reporting in a pilot are presented as a design comparison, not a demonstrated participant return. SOURCES ChatGPT Labrador News Engine — David Konitzny https://www.linkedin.com/feed/update/urn:li:activity:7520772676332761088/ Date: Reported 27 October 2026 by The Decoder; primary post relative timestamp only OpenAI and Future partner on specialist content https://openai.com/index/openai-and-future-partner-on-specialist-content/ Date: 2024-12-04 FT content partnership https://openai.com/index/content-partnership-with-financial-times/ Date: 2024-04-29 Guardian content partnership https://openai.com/index/openai-and-guardian-media-group-launch-content-partnership/ Date: 2025-02-14 Future half-year results https://futureplc.com/investor-news/future-announces-half-year-results/ Date: 2026-05-14 Tom’s Guide relaunch https://futureplc.com/blog/toms-guide-relaunches-with-smarter-shopping-tools-expert-qas-and-expanded-video-strategy/ Date: 2026-04-15 The Decoder report of Konitzny post https://the-decoder.com/openais-partner-content-arrives-in-chatgpt-pre-summarized-while-the-rest-of-the-web-gets-raw-scraps/ Date: 2026-10-27 Update on AI licensing from PLS CEO Tom West https://pls.org.uk/news-events-policy/news/update-on-ai-licensing-from-pls-ceo-tom-west/ Date: October2026, as listed by PLS news index; exact day unstated Google search publisher conduct requirement https://www.gov.uk/find-digital-markets-measures/google-search-publisher-conduct-requirement Date: 2026-06-03, updated2026-08-04 Publishers and Developers FAQ https://help.openai.com/en/articles/12627856-publishers-and-developers-faq Date: Undatedrollingguidance checked28Oct2026 Additional checked primary source: Microsoft Advertising, Tim Frank, 3 February 2026 — Building Toward a Sustainable Content Economy for the Agentic Web https://www.about.ads.microsoft.com/en/blog/post/february-2026/building-toward-a-sustainable-content-economy-for-the-agentic-web The announcement describes publisher-defined terms, usage-based reporting, voluntary participation, selected Copilot scenarios and expansion of a pilot. It does not establish a quantified payout to a named participant or affiliate commission preservation.