OpenAI accounts for 670 of 1,377 public references linking UK companies to an AI model provider, or 48.7%, according to a research report by GreenPT dated 3 November. Anthropic follows with 322 references, or 23.4%, open-source models with 235 (17.1%) and Google with 150 (10.9%). The report puts OpenAI at 2.1 times Anthropic’s count.
These are shares of references, not of companies or spending. GreenPT says a company “is counted once per provider, so these are shares of references, not of companies”. By our arithmetic, OpenAI’s lead over Anthropic is 25.3 percentage points of references.
Business Pill · Concentration Explained
A short explainer on market concentration. It teaches the general idea only and does not discuss this story.
The key insight: As we read it, GreenPT’s count puts OpenAI at 48.7% of 1,377 public references linking UK companies to a model provider and Anthropic at 23.4%; by the report’s own estimate about 31% of references to the named providers are offers, pilots, plans or stale statements rather than corroborated use.
GreenPT says it collected public evidence of AI model use by UK companies: case studies, supplier disclosures, privacy notices and processor lists, and company statements. Each reference links one company to one model provider. The report sets two reading rules: “References overlap: a company using OpenAI and Anthropic appears under both, so shares are not company counts or market shares”, and “the sample is companies with public evidence, not a representative sample of UK businesses, so no national adoption rate follows from it”.
AI Insider, which reported the study on 6 November, describes GreenPT as “a European AI platform that markets services focused on privacy and sustainability”. GreenPT’s own site describes it as “Sustainable AI powered by renewable energy and open-source models”.
Share of 1,377 public references linking a UK company to an AI model provider, from GreenPT’s report of 3 November 2026. A company that names two providers is counted under both, so these are shares of references, not of companies, customers or spending.
How Much of It Is Actual Use
GreenPT coded a second sample of 734 references across 15 providers. It found corroborated use in 69.9% of OpenAI references, 68.5% of Anthropic references and 65.7% of Google references, and says the gaps are within sampling noise. About 31% of references to the three named providers, it says, are “conditional or stale statements, offers, pilots or plans”.
All 235 open-source references were coded as corroborated use. Applying the sample rates to the headline totals, GreenPT estimates that 1,023 of the 1,377 references, or 74.3%, are corroborated use. The report labels those counts as derived: “The derived counts are an assumption, not a measurement”.
Where Anthropic Leads
In GreenPT’s industry cut, which uses a different base of 969 references, OpenAI leads the named providers in most sector groups. Anthropic leads in retail, food and hospitality, with 16 corroborated references against OpenAI’s 10, and ties OpenAI in marketing, media and creative and in healthcare, the report says. Open-source models hold 28% to 33% of corroborated use in every one of the nine sector groups.
Software, IT and AI services account for 392 of the 969 references, about 40%. Education has the highest share of corroborated use, 93.7% (74 of 79), and marketing, media and creative the lowest, 63.5% (47 of 74).
Other Counts Rank Providers Differently
Different sources put different providers first. A study of 9,821 UK annual reports by Bart Jaworski and Andrew Sutton, posted on arXiv on 1 October, says named vendor mentions in those reports “cluster around a small set of major providers led by Microsoft”. GreenPT’s headline ranking counts model providers and does not list Microsoft.
On consumer spending, a working paper covered in our 5 November piece put Claude at 29.4% of direct UK spending on AI subscriptions in July 2026 and ChatGPT at 52.5%. That measures money from bank and card records, not public references, so the figures are not comparable with GreenPT’s; we set them side by side only to show that the two counts place Anthropic in second place.
As we read the report, it measures who UK companies name in public, not who they pay: a company naming two providers counts twice, and companies that disclose nothing are absent.
As we read the coded sample, the three named providers have similar rates of corroborated use, 65.7% to 69.9%, so OpenAI’s lead in the count comes from the number of references, not from stronger evidence behind them, which is also how GreenPT describes it.
As we read the three sources together, the ranking depends on what is counted: GreenPT’s references put OpenAI first, an annual-report study says vendor mentions are led by Microsoft, and a spending study placed Claude second behind ChatGPT.
GreenPT Research report, 3 November 2026
“Corroboration does not separate the big three”
Three Implications
FOR OPENAI In GreenPT’s count it has 670 references, 2.1 times Anthropic’s 322, the report says.
FOR ANTHROPIC GreenPT says Anthropic leads in retail, food and hospitality and ties OpenAI in two other sector groups.
FOR OPEN-MODEL PROVIDERS GreenPT says open-source models hold 28% to 33% of corroborated use in each of nine sector groups.
The framework says: “Every technology stack has layers. Some are scarce, some are abundant.”
As we read it, GreenPT’s count shows the model layer as UK companies name it in public: two closed providers hold 72.0% of references, and open-source models 17.1%. The count does not show spending or which layer is scarce. The framework does not discuss GreenPT, UK companies or this report.
What Is Not Established
GreenPT’s shares are not market shares, customer counts or spending, and the sample covers only companies with public evidence. The corroborated-use counts for OpenAI, Anthropic and Google are projections from a smaller coded sample, by GreenPT’s own description. GreenPT counts open-source models as one row, while the others are single companies. We did not contact GreenPT, OpenAI or Anthropic.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
In GreenPT’s count of 1,377 public references, OpenAI is named in 48.7% and Anthropic in 23.4%, together 72.0%. About 31% of references to the named providers are not corroborated use, by the report’s estimate, and the count measures public mentions rather than adoption.
A note on sourcing. The counts, shares and corroboration rates are from GreenPT’s research report of 3 November 2026, which describes its derived counts as projections. The description of GreenPT is from AI Insider’s report of 6 November and GreenPT’s own site. The annual-report comparison is from the abstract of arXiv paper 2610.02281. The consumer-spending figures are from our 5 November piece. The 25.3-point gap is our arithmetic. We did not contact GreenPT, OpenAI or Anthropic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
