AI stocks tumble as Russia's medical AI market generates just 1.5 billion rubles

The AI market is in turmoil: after November 5, technology stocks lost around $2.5 trillion in a single trading session, Nvidia fell 6–9%, and Micron and AMD dropped 15–18%. Shortly beforehand, Ray Dalio had called AI a "classic bubble." But for a Russian clinic, overseas companies' share prices are not a basis for making decisions. A different question matters more: how to distinguish real benefits from inflated expectations.

What US reports show

Bessemer and Bain surveyed 226 healthcare executives about 65 AI use cases. Participants had budgeted for a payback period of around 24 months. By their own accounts, however, their investments paid off in roughly 12 months, with returns reaching 3.5 times costs. In 54% of use cases, respondents saw a substantial impact within the first year.

There is an important caveat: these are self-assessments, not a clinical trial or an audit. An average across different use cases does not guarantee that your project will return $3.5 for every dollar invested.

Administrative AI has an advantage because its impact is easier to measure: the number of requests processed, errors corrected, or minutes spent on a task. According to Bessemer, around two-thirds of revenue cycle solutions use semiautonomous or autonomous agents, compared with roughly 4% in clinical work. Where results can be verified, organizations are willing to place greater trust in the system.

Doctors remain cautious, however. Among respondents, 46% expressed trust in AI for clinical decisions, 77% reported that recommendations were rejected more than half the time, and 48% of insurers were unwilling to pay for fully autonomous care. The implication: tools that assist doctors, rather than replace them, have the advantage.

What makes Russia different

According to the "Zdrav AI 2025" ranking, specialized developers' actual revenue totals up to 1.5 billion rubles. Estimates of the potential market size (64.4 billion rubles) and forecasts for 2030 (78–121 billion rubles) are different types of figures and should not be placed in the same column. The government is both the main customer and the regulator, so revenue at leading companies—Webiomed, Tretye Mnenie and Celsus—depends on decisions by the Ministry of Health and the compulsory health insurance system, known as OMS.

The company figures are revealing. SberZdorovye's income grew to more than 12.8 billion rubles, but it is an ecosystem, not a pure-play AI developer. Tretye Mnenie brought in more than 270 million rubles (+79.8%), including grants and subsidies. Webiomed's income fell by approximately 30% in 2025, to 105 million rubles, against expenses of 183 million. Celsus reported income of 135 million rubles against expenses of 177 million.

For now, success favors companies with traffic, an established customer base and their own models. Specialized teams are growing severalfold, but from a low base and largely on government funding. Investment in Russian medical AI in 2018–2025 totaled around 4.7 billion rubles across 264 deals, including grants.

What to do in practice

The author proposes three tests for any project. First: is there a measurable outcome in rubles, minutes or clinical indicators, and has a baseline been recorded? Second: who pays after the pilot? Third: who will maintain the system a year from now?

For product managers and founders: build around a clear payment channel—an OMS reimbursement rate, a regional program or a clinic network's budget. Allow for lengthy approval cycles, compare market "potential" with actual revenue, and examine the share coming from grants separately.

For a private clinic, the most realistic near-term opportunities are documentation, patient routing, contact centers, front-office operations and revenue cycle management. Clinics can calculate the impact themselves. But this is the author's hypothesis: there are no publicly available Russian data on returns from administrative AI.

The conclusion: falling stock prices will not stop medical AI in Russia, because it already depends on budgets and reimbursement rates rather than the stock market. The greater risk lies in slides touting market potential without revenue.