How Claude Took Over the Enterprise AI Conversation

At the HumanX AI conference in San Francisco last week, 6,500 executives, founders, and investors gathered at the Moscone Center to talk about where enterprise AI is heading. The question on most lips was not which OpenAI product to deploy next. It was which version of Claude.

That is a meaningful shift. Twelve months ago, at a comparable industry gathering in Las Vegas, OpenAI dominated the conversation. Products, pricing, partnerships: ChatGPT was the reference point everyone oriented around. This year, attendees and observers across outlets including TechCrunch, CNBC, and Bloomberg noted the same thing independently: Anthropic has taken over the mindshare.

The Numbers Behind the Shift

Anthropic arrived at HumanX with a number worth taking seriously: its run-rate revenue has surpassed $30 billion annually, according to statements made around the conference. The company is currently valued at $380 billion. That figure would have seemed speculative eighteen months ago, when Anthropic was still widely regarded as a safety-focused research lab that happened to ship a chatbot.

Those numbers matter for reasons beyond investor interest. Enterprise buyers pay attention to vendor viability. A supplier at $30 billion in annual revenue is not going away. It has the resources to maintain product development, honor contracts, and build integrations. For large organizations evaluating multi-year AI commitments, that stability calculus has shifted noticeably toward Anthropic.

Claude Code and the Mania Signal

The specific product driving the conversation at HumanX was Claude Code, Anthropic's coding agent. Arvind Jain, CEO of enterprise AI company Glean, described the phenomenon with unusual candor: "It has become a religion, that's the level of that mania." His framing was not marketing. It was a business leader describing real pressure from employees and peers to adopt and deploy Claude Code across development workflows.

That kind of bottom-up pressure is a meaningful leading indicator. Enterprise software adoption rarely starts with procurement; it starts with teams using a tool that works unusually well and building a case for broader deployment. The pattern is familiar: how Slack entered enterprises, how GitHub Copilot spread through engineering organizations, and now how Claude Code is moving from individual developer enthusiasts to executive decision points.

What makes Claude Code specifically compelling for engineering teams comes down to a few practical factors: context window size (Anthropic's models have consistently led on usable context length), performance on software engineering benchmarks, and a reputation for following complex multi-step instructions reliably. Whether that reputation is fully deserved is debatable. In enterprise software, perception moves purchasing decisions, and right now the perception favors Claude.

OpenAI's Position

None of this means OpenAI is losing. ChatGPT remains the largest consumer AI product by usage and name recognition. OpenAI launched GPT-5.4 Thinking in April 2026, its most capable reasoning model to date, alongside a new $100 per month Pro plan filling the gap between its $20 Plus tier and $200 per month maximum plan. The company is still the default AI choice for most individual users.

But "default for consumers" and "preferred for enterprise development" are different markets with different buying criteria. Consumer products win on simplicity, familiarity, and breadth of features. Enterprise tools win on reliability, API quality, context handling, and fit for specific technical workflows. In that second market, the conference signals suggest Anthropic is no longer the scrappy challenger. It is the tool that enterprise developers are actively advocating for internally.

What the Conference Signal Is and Is Not

It is worth being precise about what "winning mindshare at a conference" actually measures. HumanX drew a particular slice of the industry: sophisticated enterprise buyers, investors, and founders with a bias toward technical deployment. That population is not representative of the broader AI market, which includes millions of individual users for whom OpenAI's consumer products remain the primary touchpoint.

The HumanX signal is a leading indicator for enterprise procurement, not a measure of overall market share. It suggests where medium and large organizations are directing their AI budgets over the next twelve to eighteen months. It does not tell you what tool a student or small business owner should use today.

Practical Takeaways for Teams Evaluating AI Tools

If you are making a decision about AI tooling for a development team or enterprise workflow, the HumanX signal is relevant but not determinative. Here is what the data actually suggests.

For engineering and coding workflows, the case for trialing Claude is strong enough that ignoring it means potentially missing the tool your engineering team will request six months from now. The adoption pressure Jain described at Glean is not unique to Glean. Our comparison of ChatGPT, Claude, and Gemini covers the specific capability differences across the leading models.

For general business use cases — writing, summarization, customer-facing chatbots, document processing — the gap between the major models is narrower. Factors like existing integrations, pricing tiers, and team familiarity carry more weight than benchmark rankings. The best AI chatbots of 2026 guide covers the full range of options across different use cases.

For teams already invested in OpenAI's API, switching costs are real. GPT-5.4 Thinking is a capable model with genuine improvements in complex reasoning tasks. The case for switching solely because of conference sentiment is weak. The case for running a structured evaluation is not.

The AI model market in 2026 is not a winner-take-all situation. OpenAI, Anthropic, and Google are all shipping capable models on aggressive release cadences. What the HumanX conference measured was something more specific: among the enterprise professionals who pay close attention to the leading edge, Anthropic has earned a level of trust that translates into deployment decisions. That took eighteen months. It is worth understanding why.