Card transaction tracking of about 28 million U.S. consumers shows Anthropic's paying base up roughly 75% since January 2026, and the lift didn't reverse in the weeks after the company refused to license Claude for mass surveillance and autonomous weapons.
For most of the past three years, Anthropic has been cast as the AI lab developers and enterprise buyers reach for, while OpenAI's ChatGPT owned the consumer subscription market. New card-transaction data suggests that split is shifting, and the shift lines up with a specific choice Anthropic made in March 2026: the company publicly refused to let its models be used by the Trump administration for mass surveillance of Americans or autonomous weapons work. Six weeks after that refusal, Indagari's payments panel still shows Claude paying subscribers growing.
The dataset comes from Indagari, a consumer-payments analytics firm that tracks roughly 28 million U.S. consumers through anonymized credit card transactions, with weekly readings running from 2025 through May 10, 2026. Anthropic's paid-consumer count and revenue inside that panel are up about 75% since January 2026, according to TechCrunch's reporting on the Indagari figures. Critically, the growth continued past the March spike rather than fading with the news cycle, which is what makes it worth reading as more than a one-week protest bump.
The mechanism behind the lift is what makes the story durable. Anthropic's refusal was not a press-release abstraction; it named specific government uses and drew a line. An earlier TechCrunch piece from March 28 had already flagged Claude's rising paying-consumer share, which means the March refusal landed on top of an existing upward trend rather than starting one. For consumers weighing which AI subscription to keep month after month, a public ethical stance by the company that builds the model appears to be functioning as a brand cue that translates directly into recurring payments.
Independent data lines up with the directional reading. Sensor Tower's State of AI 2026 report projects global time spent on generative-AI apps will more than double year-over-year. App usage and card swipes are different signals, but read together they suggest the consumer AI market is expanding while Claude is taking share inside it, not just riding a rising tide.
Three limits matter for how far to push this. Indagari's panel is a sample of U.S. card transactions, not Anthropic's full revenue or total customer count, and the methodology is anonymized and trend-oriented. The dataset ends May 10, 2026, so any claim about the post-May picture is, by definition, roughly six weeks stale. And the March refusal event is itself sourced through TechCrunch's narrative rather than a primary Anthropic filing or earnings call, so the alignment between the refusal and the payments lift is an inference, not a controlled experiment.
The next data point is whether the lift holds. If Indagari's June and July readings show the 75% growth cooling, flattening, or reversing once they land, the story becomes a durable consumer trend that happened to crest near a high-profile refusal. If the growth keeps extending, Anthropic will have turned one policy decision into the clearest evidence yet that an explicit ethics stance can move the consumer AI subscription market.