Why consumer AI's popularity surge masks a persistent profitability crisis

Consumer AI is having a visible moment. Meta's Muse personal assistant and OpenAI's newly launched Dots avatar are both riding a wave of user interest, while the viral assistant Instinct recently hit a $10 billion valuation on the back of its ability to handle everyday errands. To outside observers, the convergence looks like ChatGPT's 2022 breakthrough moment all over again — raw AI capability finally meeting consumer demand.

Yet beneath the enthusiasm lies a fundamental problem that has already prompted the industry's largest players to dramatically shift strategy. The brutal reality: consumers simply aren't willing to pay enough to make consumer AI businesses profitable, and that willingness shows little sign of improving even as underlying models become dramatically more powerful.

The bleak numbers#

Recent market data tells a cautionary tale. According to Andreessen Horowitz's latest State of Markets report, drawing on PNC research from summer 2026, only 2.2% of consumers paid for AI services as of May, with an average monthly spend of $31. The growth trajectory appears disturbingly linear — even enormous performance leaps between successive model generations barely register on adoption or spending charts.

Bank of America and other analysts offer slightly higher estimates, with roughly 3% of U.S. consumers paying for AI-related services. A more optimistic Menlo survey suggests a quarter of adults use AI daily, though this includes unpaid offerings. Either way, the ceiling remains low.

The math becomes stark when compared against established benchmarks. Netflix operates with 325 million global subscribers and generates roughly $11 billion annually per dollar of ARPU. At the current $31-per-month figure, a fully saturated consumer AI market would generate annual revenue well below $11 billion — far short of what OpenAI alone reportedly spends to operate.

The cost problem nobody talks about#

Revenue constraints aren't the only issue. AI services carry infrastructure and computational expenses that dwarf those of earlier consumer tech waves. Social networks and cloud platforms can operate at scale with relatively modest per-user costs; AI systems require continuous, expensive inference and potential fine-tuning. That fundamental difference means profitability remains elusive even if user acquisition accelerates dramatically.