Anthropic Dominates Enterprise AI Adoption as Microsoft Maintains Market Lead
Anthropic is hemorrhaging into the enterprise faster than anyone else in the AI tooling race — and that's not hyperbole, it's the math.
According to a new analysis by identity management firm Okta, drawn from anonymised access data across more than 20,000 companies over four years, Anthropic's corporate account growth has outstripped every major competitor by a staggering margin. Yet the bigger story isn't who's winning the sprint. It's that the enterprise AI market has quietly settled into a shape nobody predicted: not a winner-take-all monopoly, but a sprawling, multi-platform patchwork — and the companies stitching it together are moving faster than the vendors chasing them.
The numbers behind the surge
Okta tracked net increases in corporate accounts from June 2022 through June 2026, setting Anthropic's growth rate as a baseline index of 100. OpenAI landed at 66.9. Google Workspace at 59.8. GitHub at 56.6. Cursor — the AI-native code editor that's become a developer darling — trailed at 42.4. The survey scoped knowledge-worker and developer tools: foundation models, enterprise search, collaboration suites, meeting AI, creative product platforms. What it excluded — cloud-infrastructure-delivered AI, cybersecurity-specific tools, department-level work systems — means the picture isn't exhaustive, but it's the clearest window yet into where corporate budgets are actually flowing.
The kicker: Anthropic's growth didn't just lead — it accelerated. After January 2026, the company's corporate account numbers surged while OpenAI and Cursor's pace moderated. Okta flagged Anthropic separately as a service experiencing "a notable recent influx" of business customers. That kind of inflection point, four years into a market, signals something deeper than hype. Enterprises aren't experimenting anymore. They're committing.
Installed base still belongs to the old guard
Growth velocity and total footprint are two different animals. Microsoft 365 remains the undisputed heavyweight by sheer number of corporate accounts, with Google Workspace in second and aggressively narrowing the gap. Anthropic's adoption speed is extraordinary, but the gravitational pull of entrenched productivity platforms — the ones already woven into compliance frameworks, procurement cycles, and a thousand Slack integrations — shouldn't be underestimated.
This tension between momentum and incumbency is playing out across the entire tech stack. Traditional financial institutions, for instance, are racing to build infrastructure around emerging technologies at a pace that would have seemed absurd five years ago — a dynamic mirrored in moves like Standard Chartered's first digital asset prime brokerage trades with LMAX, where legacy banking muscle meets nascent market infrastructure. In enterprise AI, the same pattern holds: the newcomers are winning attention, but the incumbents own the pipes.
No single vendor — and that's the point
Perhaps the most telling data point: roughly 57 percent of surveyed companies now run at least two AI platforms simultaneously. Enterprises have embraced what Okta calls a "right tool for the right job" strategy — layering Claude for one workflow, GPT for another, Cursor for code, Google's suite for search and collaboration. They're not pledging allegiance. They're assembling toolkits.
This is the market structure that should keep every AI vendor up at night. Not a zero-sum contest for a single throne, but a competitive landscape where differentiation by task, integration depth, and pricing elasticity matters more than brand loyalty. Microsoft and Google are leveraging their installed bases to cross-sell AI into existing workflows. Anthropic is surging by winning net-new adoption in roles where its models demonstrably outperform. OpenAI and Cursor remain top-tier but are watching their relative growth cool.
For anyone making procurement decisions or building on these platforms, the takeaway is blunt: multi-platform fluency isn't optional anymore. The companies adopting fastest aren't picking winners — they're hedging, testing, and letting workflow-level performance dictate where each tool earns its seat. The enterprise AI market hasn't consolidated. It's diversified. And the pace of that diversification is the real story Okta's data is telling.