TechCrunch's latest analysis of enterprise AI usage carries a blunt message for anyone who assumed Anthropic had the business market sewn up: OpenAI is closing the gap fast. For UK founders, CTOs and operations leads who backed Claude for coding and enterprise reliability, the practical takeaway isn't 'pick one and forget it' — it's 'match the model to the job, and revisit that decision every quarter'.
What is the Concept
OpenAI and Anthropic have taken different routes to the same customers. OpenAI built its business base off the back of ChatGPT's consumer scale, then pushed ChatGPT Enterprise and its API into finance, retail and professional services. Anthropic grew more slowly but earned a reputation among engineering teams for Claude's coding ability and its emphasis on predictable, safety-conscious outputs — which made it the default choice for regulated or risk-averse organisations.
'Gaining on' in this context means seat growth, API spend and enterprise contract wins — not just chatbot popularity. When TechCrunch reports new data showing OpenAI narrowing Anthropic's lead with business users, it signals that procurement teams who previously defaulted to Claude are now genuinely comparing both vendors before signing.
Why It Matters in United Kingdom (2025–2026 Context)
UK businesses are under sustained pressure to justify every software line item. With interest rates still elevated and budgets tight across London, Manchester, Leeds and Edinburgh's tech clusters, finance directors are scrutinising AI subscriptions the same way they scrutinise cloud hosting bills. A genuine two-horse race between OpenAI and Anthropic is good news here: it gives UK buyers negotiating leverage that didn't exist eighteen months ago, when Anthropic could effectively set enterprise pricing without much resistance.
This also lands at a moment when UK procurement teams are formalising AI vendor assessment as part of standard due diligence, partly driven by ICO guidance on data processing and partly by boards wanting documented reasoning behind software spend. A company that can show it evaluated both providers — and picked based on evidence rather than habit — is in a stronger position when auditors or investors ask how AI decisions get made.
How AI Is Changing This
The competitive front line has moved to coding and agentic workflows. Anthropic's Claude Code pulled ahead early with developers, but OpenAI has closed much of that gap with its own coding-focused releases and deeper IDE integrations. Both vendors are now shipping agent frameworks that can execute multi-step business tasks — invoicing, customer support triage, contract review — rather than just answering questions, which shifts the buying decision from 'which chatbot is smarter' to 'which agent stack integrates with our existing tools with the least engineering effort'.
The practical result for UK teams is that a single-vendor default is becoming harder to justify. Mid-sized UK software teams are increasingly running OpenAI models for customer-facing and general-purpose tasks while keeping Anthropic models for code review or compliance-sensitive drafting — treating the two as complementary tools rather than competitors to choose between once and never revisit.
Real-World Examples
Consider a Manchester-based fintech scale-up processing loan applications. Its engineering team standardised on Claude for code generation because of its lower error rate on financial logic, but switched customer support drafting to OpenAI's API after finding it cheaper per token at similar quality for that specific use case. Neither decision was ideological — both were driven by a monthly cost-and-accuracy review, which is the discipline more UK businesses are now adopting instead of locking into a single annual enterprise contract.
This is exactly the kind of decision RP SoftTech helps UK businesses navigate — running structured, side-by-side pilots across OpenAI and Anthropic models against a company's actual workflows before recommending an integration, rather than defaulting to whichever vendor has the loudest marketing that quarter.
Practical Insights / Actions
Use what we'd call the Vendor Leverage Window framework: while two credible providers are actively competing for the same enterprise customers, buyers hold pricing and contract power they won't have once the market consolidates. UK businesses should treat 2026 as inside that window and negotiate accordingly — shorter contract terms, usage-based pricing options, and explicit rights to run parallel trials before renewal.
The most common founder mistake right now is signing a 12-month enterprise agreement with either vendor based on a single team's preference, without running a documented comparison against the actual task mix the business needs — support, coding, analysis, content. The hidden opportunity is a multi-model stack: routing each task type to whichever model performs best and costs least for it, which typically cuts blended AI spend by a meaningful margin compared with an all-in-one-vendor contract, while also reducing the business risk of depending on a single provider's uptime and pricing decisions.
Future Outlook
Expect the OpenAI-Anthropic rivalry to keep pushing prices down and capability up through 2026, which benefits UK buyers directly. Enterprise per-seat pricing that sat firmly in the £20–£30 monthly range for advanced tiers is already showing more flexibility as both vendors chase larger UK and European contracts. The likely direction is toward orchestration layers that let businesses route tasks across both providers automatically, making the 'which one do we choose' question increasingly moot for larger organisations — though smaller UK SMEs without engineering resource to build that routing will still need to make a deliberate, reviewed choice.
The risk to watch is complacency: businesses that pick a vendor now and stop paying attention will likely overpay within twelve months, as this market is moving too fast for a 'set and forget' contract to hold its value.
Conclusion
The real winner in the OpenAI-versus-Anthropic contest isn't either vendor — it's the UK business that refuses to pick a permanent side. Treat 2026 as a year for structured evaluation, shorter contracts and task-based model selection, and use the current competitive pressure to negotiate better terms than either provider would offer in a less contested market.