Artificial intelligence has finally ceased to be exotic and has firmly entered the everyday financial habits of Russians. My analysis of payment infrastructure data for July shows impressive dynamics: the share of cardholders paying for subscriptions to AI services has reached 51.6%. In other words, every second user paying by card now has at least one active subscription to a neural network.
The average monthly spend on such services is 2,097 rubles. Notably, a one-time payment typically equals 1,905 rubles — exactly the cost of the basic tariff of a popular chatbot, including taxes. On average, users make two such payments per month, which indicates the formation of a stable habit rather than one-off experiments.
One service instead of a dozen subscriptions
The widespread belief that users accumulate many AI subscriptions is not confirmed. The data shows the opposite: 74.3% of payers pay for only one product, while another 23.9% pay for two or three. Only 0.1% of users have seven or more active services, but it is this microscopic group that accounts for about 1% of all AI spending.
On average, there are 1.35 services per person, and this figure has been stable for six months. Users find a suitable tool and work with it rather than collecting dozens of products. Of particular interest is the group of "heavy" users: 15.2% of payers spend more than 10,000 rubles per month, providing 57% of the entire market turnover. For them, AI is no longer a subscription for convenience but a full-fledged work tool, which is why they prefer pay-as-you-go over fixed tariffs.
Turnover grew nearly fourfold in five months
From March to July, turnover on AI services increased 3.85 times, and the number of paying users grew 3.45 times. At the same time, average spending per person remained almost unchanged: 2,178 rubles in March versus 2,097 rubles in July. This is a clear signal: the market is growing in breadth due to an influx of new users, not through increased spending by existing ones. AI is moving into the category of ordinary monthly expenses — similar to streaming services.
Experts attribute this dynamic to neural networks becoming a familiar line item in the budget. However, it is important to understand: for businesses, demand is growing even faster. On the Yandex AI Studio platform, clients spent 597 billion tokens in the first half of the year — 18 times more than a year earlier, with Chinese models Qwen, GLM, and Kimi leading the way.
My comment: The AI market in Russia is going through a stage of mass adoption, which will inevitably affect adjacent sectors, including mining infrastructure and data centers. The shortage of computing power and dependence on imported GPUs will become key growth constraints. While businesses and the state seek workarounds — from tax incentives to testing Chinese chips — the retail user will continue to vote with their rubles for convenience and functionality, making this market one of the most promising for long-term investment.