The ugly economics of consumer AI

Consumer artificial intelligence is experiencing a renewed wave of interest driven by personal assistants like Meta’s Muse, OpenAI’s Dots, and the up-and-coming Instinct assistant, which recently reached a $10 billion valuation. Despite reliable agentic capabilities handling everyday tasks like booking travel and managing subscriptions, the sector faces challenging economics.

According to an Andreessen Horowitz report citing summer PNC research data, only 2.2% of consumers paid for AI services as of May, with an average monthly spend of $31. Bank of America reported in March that roughly 3% of U.S. consumers paid for AI, while a September Menlo survey indicated that a quarter of adults use AI daily, with half of those users paying.

However, operating costs remain exceptionally high compared to previous technologies, and consumer spending falls short of covering expenses. Consequently, frontier labs and companies like OpenAI have shifted focus toward enterprise contracts. While products like Muse and Instinct pursue different monetization paths—such as Meta’s ad targeting and Instinct taking a cut of purchases—underlying consumer AI economics continue to limit growth without enterprise revenue.