Whether you're looking for an angel investor, a growth advisor, or just want to connect — I'm always open to great ideas.
Get in TouchNew credit card transaction analysis of 28 million US consumers confirms Claude's paid subscriptions more than doubled since January 2026, driven by viral Pentagon drama, Super Bowl ads, and Claude Code buzz — and the data suggests it's pulling users directly from ChatGPT.
AI, startups & growth insights. No spam.
For the first time, hard financial data — not self-reported metrics or survey estimates — is confirming what many in the AI industry had suspected: Claude is eating into ChatGPT's paying user base at an accelerating pace. An analysis of billions of anonymized credit card transactions from roughly 28 million US consumers, commissioned by TechCrunch and conducted by consumer transaction intelligence firm Indagari, shows that Claude's paid subscriptions more than doubled between January and late March 2026. The growth isn't just fast — it's accelerating. And the forces driving it, from a headline-grabbing Pentagon standoff to Super Bowl TV spots mocking OpenAI, are unlike anything the AI industry has produced before.
The most important thing to understand about this analysis is that it measures revealed preference — actual payment behavior — not stated preference. When someone pulls out a credit card and pays $20 a month for Claude Pro, that transaction gets captured. Surveys can be gamed. App downloads inflate vanity metrics. But recurring subscription charges on a credit card are an unambiguous signal of sustained demand.
Indagari, the consumer transaction intelligence firm that conducted the analysis for TechCrunch, examined billions of individual credit card transactions from approximately 28 million anonymized US consumer accounts. The methodology looks for charges from Anthropic's billing entity on recurring monthly cycles — a reliable proxy for active Claude Pro subscribers.
The headline finding: Claude's paid subscription count more than doubled from January 2026 through late March. Anthropic confirmed the figure to TechCrunch, adding that daily active users have more than tripled over the same period.
The majority of new subscribers are joining at the base "Pro" tier — $20 per month. That price point puts Claude Pro in direct competition with ChatGPT Plus, which is also priced at $20 per month. The overlap is not coincidental. It is the battleground.
What the data cannot tell us is the exact total number of Claude subscribers, since the 28 million consumer panel doesn't represent every US consumer. Estimates for Claude's total user base range widely — from 18 million to 30 million across all tiers — and Anthropic has not published official subscriber counts. What the data does confirm is the direction and magnitude of the trend: sharply upward, at a pace that has accelerated through Q1 2026.
The most counterintuitive chapter of Claude's subscriber story begins in late January 2026, when Anthropic's CEO Dario Amodei drew a line in the sand with the US Department of Defense. Amodei publicly refused to allow Claude to be used for lethal autonomous weapons systems or for mass surveillance of American citizens — uses the DoD had reportedly sought to include in its AI contracts.
What followed was a cascade of escalating confrontations. On February 24, Defense Secretary Pete Hegseth issued an ultimatum: relent by 5:01 p.m. on Friday, February 27, or face consequences. Amodei did not relent. On February 27, President Trump directed federal agencies to cease using Anthropic products, and the Pentagon formally designated Anthropic a "supply chain risk" — an extraordinary measure typically reserved for adversarial foreign technology suppliers.
The backlash was immediate. Anthropic filed suit. A federal judge in California issued a preliminary injunction on March 26, blocking the Pentagon's designation and writing in a 43-page ruling that "nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary and saboteur of the U.S. for expressing disagreement with the government." (See our earlier coverage: Anthropic Wins Injunction Against Trump Pentagon DoD Ban.)
For a subscriber growth story, you might expect this kind of government confrontation to be a disaster. Instead, the opposite happened. Claude climbed to No. 1 on the App Store on March 1 — the day after Trump's ban took effect. The Indagari data shows a measurable subscription spike in late February and early March that directly correlates with the Pentagon coverage cycle.
The explanation is not complicated: in a political environment where many consumers had grown skeptical of major tech companies capitulating to government pressure, Anthropic's visible refusal to comply resonated. Claude was suddenly a brand with a story, not just another AI chatbot. Consumers who had been vaguely aware of Claude converted to paid subscribers at elevated rates. Return subscribers — previous users who had lapsed — also came back in record numbers during February, per Indagari's data.
The Pentagon story was reactive. The Super Bowl campaign was deliberate — and it worked even better.
Anthropic aired several commercials during Super Bowl LX in early February featuring darkly comedic scenarios that explicitly mocked ChatGPT's decision to introduce advertising to its free tier. The core message was blunt: Claude will never show you ads. The ads were unusual in their directness — most enterprise software brands play it safe during major broadcast events. Anthropic leaned in.
The results were swift. According to CNBC, Anthropic saw an 11% user boost in the period immediately following the Super Bowl broadcast. Claude's app jumped from No. 41 in the US App Store to a top-10 position within days of the ads airing. Website traffic spiked 6.5% in the post-Super Bowl window.
Sam Altman, OpenAI's CEO, was notably agitated in public about the campaign. His reaction was itself a news cycle — which amplified the Anthropic ads' reach beyond the initial broadcast audience at no additional cost to Anthropic.
The timing of the Super Bowl ads relative to the Pentagon dispute created an unusual compounding effect. Between early February and early March, Claude had two separate major news narratives driving consumer awareness: the "anti-ads" positioning versus OpenAI, and the "standing up to government pressure" narrative from the Pentagon fallout. The Indagari data shows subscription growth accelerating through both events, with the Pentagon spike in late February arriving on top of an already-elevated baseline from the Super Bowl effect.
The Pentagon story and the Super Bowl ads drove mainstream consumer awareness. But a quieter, more durable growth driver was also at work during the same period: Claude's developer and productivity tools.
Anthropic released Claude Code and Claude Cowork in January 2026. Claude Code is a developer-focused product that allows programmers to use Claude directly from their terminal and IDE environments, handling complex multi-file coding tasks with minimal supervision. Claude Cowork is an enterprise collaboration tool built around Claude's reasoning capabilities.
The developer community's response to Claude Code was intense. The product generated significant organic word-of-mouth on platforms like X (formerly Twitter) and Hacker News, with engineers sharing benchmarks and workflow demonstrations. A widely-cited Medium piece titled "Claude Code — the best $20 I've spent in a while" captured a sentiment that spread through software developer networks, directly driving subscription conversions at the Pro tier.
Anthropic confirmed to TechCrunch that Claude Code has been a meaningful driver of paid subscription growth throughout Q1 2026. Developer subscribers tend to convert at higher rates than general consumers and churn at lower rates — they embed the tool into their professional workflows and maintain subscriptions regardless of competing product launches.
Then, in late March, Anthropic released Computer Use — a capability that allows Claude to navigate a computer independently, clicking, scrolling, and taking actions on its own without requiring explicit step-by-step instructions. Per Anthropic's statement to TechCrunch, the Computer Use release drove another measurable surge in new subscriptions in the week of launch.
The three product launches — Claude Code in January, Claude Cowork in January, and Computer Use in late March — form a product cadence that has kept subscription growth elevated even in periods between the viral news cycles of the Pentagon dispute and the Super Bowl campaign.
The credit card data confirms Claude's growth is real and substantial. It is also honest about the gap that remains.
ChatGPT has approximately 35 million paying users across its Plus subscriber tier and business accounts, according to industry analysts. Its weekly active user base reached 900 million in early 2026 — more than double the 400 million reported in February 2025. ChatGPT's daily active users on mobile alone hit 250.5 million on March 2. These are numbers Claude is not yet close to matching.
Claude's total estimated user base — including free and paid tiers — sits somewhere between 18 million and 30 million, depending on the methodology. On March 2, Claude's daily active users on iOS and Android hit 11.3 million — a figure that represented a 183% increase from the start of 2026, but still roughly one-twenty-second of ChatGPT's mobile DAU figure from the same day.
The $20 per month price point comparison does show a more competitive picture at the paying-subscriber layer. DemandSage estimates Claude had approximately 18.9 million monthly active users on its web browser app in early 2026. Claude's growth rate in paid subscriptions (more than doubling in under three months) outpaces published estimates of ChatGPT Plus's subscriber growth rate over the same period.
The interpretation of "stealing ChatGPT users" requires a nuance the credit card data alone cannot provide. Separate corporate credit card spending data from Ramp, covering more than 50,000 businesses, suggests that approximately 79% of companies paying for Anthropic also pay for OpenAI — indicating that for many customers, Claude is additive rather than substitutional. Among businesses paying for AI for the first time, however, Anthropic now wins roughly 70% of head-to-head matchups against OpenAI, a reversal from a year earlier when only one in 25 Ramp platform businesses paid for Anthropic at all.
The consumer picture may differ from the enterprise picture. For individual consumers at the $20/month tier, the economics of carrying two AI subscriptions are less compelling. The evidence that paid Claude Pro subscribers are being won partly from the ChatGPT Plus audience — rather than purely from new AI adopters — is circumstantial but consistent with the data.
Total user estimates for Claude range from 18 million to 30 million, and Anthropic has not confirmed which figure is accurate. The range itself tells a story. At 18 million, Claude is a significant but secondary player in the consumer AI market. At 30 million, it is approaching a position where it would represent a material percentage of the global AI chatbot user base.
What is not in dispute is the trajectory. Anthropic has confirmed that daily active users more than tripled from January to late March 2026 — a 3x increase in under 90 days. That kind of growth rate, sustained over a full year, would push Claude into genuinely competitive territory with ChatGPT's current user counts. Sustaining that rate is unlikely; the Pentagon and Super Bowl tailwinds will not recur indefinitely. But even at a fraction of that pace, the momentum is notable.
For context on Anthropic's broader business position: the company raised $3.5 billion in late 2025, giving it the runway to invest in consumer-facing growth even while its enterprise API business remains its primary revenue driver. The company is also reportedly weighing IPO plans for 2026, which would add additional incentive to demonstrate strong and growing consumer subscription metrics before going public.
The leaked details about Anthropic's Mythos model, described internally as a "step change" in capabilities, add a potential future catalyst. If Mythos delivers on its rumored capabilities, it would provide another marketing moment similar to the Claude Code launch — a product event that converts curious observers into paying subscribers.
Any honest reading of the Indagari data requires acknowledging what it does not measure.
The credit card transaction analysis covers consumer paying subscribers — specifically individuals with US credit cards making recurring monthly payments to Anthropic. It does not include:
Enterprise API revenue. Claude's enterprise business, which involves API access for businesses building products and services on top of Anthropic's models, is Anthropic's primary revenue source and is not reflected in the consumer credit card data. Anthropic's annualized run-rate revenue reached approximately $14 billion in February 2026 and reportedly rose further in early March, according to DemandSage analysis. The enterprise trajectory has its own dynamics, distinct from consumer subscription growth.
Free tier users. The millions of users interacting with Claude without paying Anthropic are invisible in this data. Free users represent potential future paid converts, brand awareness, and feedback for model improvement — but they do not appear in the credit card analysis.
International markets. The 28 million US consumer panel provides a strong domestic signal but cannot be directly extrapolated to Claude's global performance.
These limitations do not diminish the significance of the consumer subscription data. They do mean the full Claude growth story is larger — and more complex — than the credit card numbers alone can capture. Anthropic's enterprise business is growing in parallel, with businesses increasingly choosing Anthropic over OpenAI for new AI deployments, particularly in use cases that require strong safety and usage policy controls.
The threat Claude represents to ChatGPT is most visible at the $20/month price tier — the consumer paid subscription market that OpenAI pioneered with ChatGPT Plus in 2023.
OpenAI's decision to introduce advertising to ChatGPT's free tier — the same decision Anthropic's Super Bowl ads mocked — reflects a tension in OpenAI's business model. The company is simultaneously trying to grow its user base (via free access), monetize that base (via ads and paid upgrades), and maintain its position as the default AI choice for paying consumers. Anthropic is now competing on all three dimensions, with a consumer brand message that explicitly positions Claude as the no-ads, privacy-respecting alternative.
The 11% user boost Anthropic received from its Super Bowl ad campaign is a meaningful data point not just for Claude's growth, but for OpenAI's vulnerability. The ad worked because it identified a real and existing consumer dissatisfaction with the direction ChatGPT was taking. That dissatisfaction created a switching incentive — and the credit card data suggests a meaningful number of consumers acted on it.
OpenAI's aggregate numbers remain vastly larger. With 900 million weekly active users and a product that has become synonymous with "AI chatbot" for much of the general public, OpenAI is not in existential competitive danger from Anthropic's Q1 2026 growth numbers. But the gap in growth rates — Claude doubling its paid subscribers while ChatGPT's growth rate comparatively flattens — is the kind of trend that, if sustained, eventually reshapes market structures.
The Ramp enterprise data showing Anthropic winning 70% of new-to-AI business customers is perhaps the more strategically significant figure. Businesses that adopt Claude first are less likely to switch later. The AI market is still early enough that first-mover effects at the enterprise layer will compound over years — and those first movers are increasingly choosing Anthropic.
The TechCrunch/Indagari credit card analysis is the clearest evidence yet that Claude's growth is not a PR narrative — it is a financial reality being documented in real consumer payment behavior. Paid subscriptions more than doubled in under three months. Daily active users tripled. A combination of viral news events, deliberate marketing, and genuine product launches drove each wave of growth in sequence.
Claude is not winning the AI race. ChatGPT's user counts, brand recognition, and enterprise relationships remain in a different league. But Claude is closing the gap faster than at any previous point in its history, and it is doing so at the level that matters most for long-term business health: paying customers who have made a conscious choice to spend money on the product month after month.
The next inflection point is likely to be the Mythos model release, the Computer Use feature maturing into broader workflows, and the potential IPO timeline — all of which will generate new marketing cycles and new conversion opportunities. If Anthropic can maintain even a fraction of its Q1 2026 growth momentum through the second half of the year, the competitive map of the AI industry will look meaningfully different by 2027.
The credit cards don't lie. Claude is growing. And it's growing in the places that matter.
Sources:
a16z top 100 gen AI apps sixth edition: ChatGPT hits 900M WAU, Claude grew 200%+ paid subscribers, Gemini leads growth rate.
Apple is overhauling Siri in iOS 27 to support rival AI assistants including Claude, Gemini, Grok, and Perplexity via a new Extensions system, ending ChatGPT's exclusive role and turning iPhone into the ultimate AI battleground.
Anthropic's most powerful AI model, Claude Mythos, was accidentally exposed via an unsecured CMS database, triggering a cybersecurity stock selloff and raising urgent questions about AI safety, operational security, and what 'step-change' capabilities really mean for the industry.