Anthropic's Claude is winning over paid consumers, a market owned by ChatGPT

Anthropic's Claude is gaining ground on ChatGPT in the one metric that matters most to the AI industry's financial future: paying consumers.

New data from Indagari, which analyzes billions of anonymized credit card transactions from roughly 28 million U.S. consumers, shows Claude's paying customer base growing approximately 75 percent since January 2026. The gains continued even after the company's high-profile refusal in March to allow its models to be used by the Trump Administration for mass surveillance and autonomous weapons — a stance that might have been expected to cost it customers but instead appears to have strengthened consumer loyalty.

The data tells a nuanced story. Claude is not overtaking ChatGPT. Not even close. ChatGPT remains far and away the most popular AI tool with consumers across every metric. But Claude is gaining where it counts: in the wallets of people willing to pay for AI, and in the mindshare of users actively seeking to learn new platforms.

**The numbers behind the shift**

Indagari's transaction data covers weekly payments from 2025 through May 10, 2026, including subscriptions and API token purchases. While it cannot provide absolute revenue or customer counts, the sampling of 28 million consumers is large enough to identify meaningful trends.

Claude's trajectory among paid consumers is unambiguously upward. Month over month, the trend line moves consistently higher, with revenue from this segment up roughly 75 percent since the start of 2026.

Perhaps more telling is data from DataCamp, an online education platform with about 20 million users. "Claude" is now the most searched term on the site — more searched than "AI" itself. Among self-directed consumers, demand for Claude courses is outpacing ChatGPT courses by three to one. Demand for Claude-specific coursework has increased 18 times in the last 30 days alone.

This is a striking reversal. For most of ChatGPT's existence, it has been the default entry point for anyone learning to use AI. DataCamp's data suggests that people who are motivated enough to seek out AI education are now choosing to learn Claude first.

**The March refusal that built a brand**

In March 2026, Anthropic refused to allow its models to be used by the Trump Administration for mass surveillance of Americans and autonomous weapons systems. The decision was risky — government contracts represent a significant revenue opportunity, and refusing them could alienate federal agencies and their enormous budgets.

Instead, the refusal appears to have catalyzed consumer trust. Indagari's data shows Claude's paid consumer growth continued accelerating after the announcement, not decelerating. In a market where users are increasingly concerned about how their data is handled and how AI is deployed, Anthropic's stance became a differentiator.

This dynamic mirrors broader consumer trends. In technology markets, trust is increasingly a competitive advantage, not just a compliance requirement. Anthropic appears to have translated a principled stand into commercial momentum.

**The market ChatGPT built but doesn't fully own**

OpenAI created the consumer AI market. ChatGPT's launch in late 2022 demonstrated that millions of people would pay for AI-powered tools, and the company has since built a subscriber base that dwarfs all competitors combined.

But creating a market and owning it entirely are different propositions. The consumer AI market is still early enough that switching costs are low. Users can try multiple services simultaneously, and many do. Claude's growth among paying consumers suggests that a meaningful segment of the market sees value in having options — and that Anthropic's product is good enough to be one of them.

Sensor Tower's data confirms the gap: ChatGPT still has many more paying users than Claude across all platforms. But the gap is narrowing, and it is narrowing specifically among the consumers who generate the most revenue per user.

**What happens when both companies go public**

Both OpenAI and Anthropic are preparing for public offerings, which means the financial details that remain private today will soon be subject to SEC disclosure requirements. This is where the competitive picture will become much clearer.

Investors will want to see churn rates, revenue per user, enterprise versus consumer revenue splits, and the concentration of customer bases. The Indagari data hints at a potential narrative: Anthropic may have a more diversified revenue base than commonly assumed, with meaningful consumer income supplementing its well-known enterprise and developer business.

The unknown variable is how Anthropic's latest conflict with the U.S. government will affect its business. Earlier this month, the government banned Anthropic from allowing its most powerful cybersecurity-focused models — Mythos 5 and Fable 5 — to be used by non-Americans, prompting the company to pull them from the market entirely. The impact of this restriction on revenue and customer relationships is unclear.

**What this means for you**

If you're choosing between AI subscriptions, the data suggests both services are worth considering. ChatGPT still has the broadest ecosystem of integrations, plugins, and third-party tools. Claude appears to be winning on trust, coding assistance, and — increasingly — consumer satisfaction among paying users. A $20 monthly subscription to each gives you access to both, and many power users report that the combination is more valuable than either alone.

If you're an investor watching the AI space, the key takeaway is that the consumer market is not winner-take-all. ChatGPT built it, but Anthropic is proving it can grow within it. The forthcoming IPOs will provide the hard financial data needed to evaluate both companies' unit economics.

If you work in AI or technology more broadly, the trend is clear: consumer trust is becoming a measurable competitive advantage. Anthropic's March refusal could have been a footnote; instead, it appears to have accelerated growth. In a market defined by rapid commoditization of model capabilities, brand positioning around safety and ethics may prove to be the moat that technical performance alone cannot provide