Consumer AI Spending Has Concentrated Among Power Users

While AI adoption has grown, revenue remains tethered to a small percentage of heavy-spending subscribers.

Updated on Sept. 30, 2026 in Artificial Intelligence

Bold flat-color editorial illustration showing a large central monolith surrounded by smaller geometric data prisms, representing concentrated AI cloud computing resources.
While AI adoption has reached a majority of U.S. adults, revenue remains concentrated among a small minority of power users. AI Illustration. Upload story photo >

Live Poll

Do you plan to pay for additional artificial intelligence subscriptions in the coming months?

Data from 2026 shows that while 64% of U.S. adults have used artificial intelligence, paying subscriber bases remain small. Global consumer spending on AI reached $40 billion this year, driven largely by a minority of frequent users.

Why it matters

The shift toward AI agents, which require significantly more computing resources than standard software, places pressure on subscription business models. As companies scale, the financial viability of these services depends on whether usage broadens beyond the current concentration of heavy payers.

The top 14% of AI payers account for 60% of total consumer-AI spending, while average monthly subscription costs reached $31 in May 2026. These platforms are increasingly driven by autonomous agents like OpenAI's Dots, which can interface with over 4,000 different applications.

The players

Meta

A social media and technology company focusing on advertising-supported platforms and recent deployments of generative AI agents like Muse.

OpenAI

An artificial intelligence research and deployment laboratory that develops large language models including the GPT-6 Astra architecture.

The details

New AI agents operate via cloud computing, a model where processing tasks are performed on remote servers rather than local hardware. OpenAI's Dots, powered by the GPT-6 Astra model, function as intermediaries that execute commands across more than 4,000 distinct software applications. Meta's Muse agent similarly utilizes these server-side resources to perform personal tasks, a shift that increases compute requirements compared to traditional, static subscription software.

Timeline

  1. 2024 served as the baseline year for household AI subscription measurements.

  2. Global consumer-AI spending was estimated at $12 billion in 2025.

  3. In February 2026, 3% of Bank of America households paid for AI services.

  4. By May 2026, 2.2% of total households maintained generative-AI subscriptions.

  5. Meta and OpenAI launched new AI agent products in September 2026.

The Tech Race

This growth follows the trajectory set by the $40 billion global consumer-AI spend recorded in 2026. Companies are currently locked in a race to transition users from passive chatbot interactions to autonomous agent-based workflows that demand higher infrastructure overhead.

Users can expect AI services to become increasingly integrated with their existing software suites as agents like Dots gain the ability to manage workflows across thousands of applications. However, the move toward compute-heavy models may lead to continued subscription price volatility for consumers.

The takeaway

The gap between casual AI usage and high-volume, subscription-based interaction is widening. Readers should monitor the pricing tiers of major AI agents in the coming quarters to see if usage-based billing models replace current flat-fee subscriptions.

Further reading

For more on how shifting models affect adoption, visit our Artificial Intelligence section.

Source note: This article includes information reported by TokenPost.

Live Poll

Do you plan to pay for additional artificial intelligence subscriptions in the coming months?