Really…. Will Consumers Ever Pay for AI?
AI is everywhere.
We use it to write, search, create images, edit videos, plan trips, write code and increasingly, get things done.
But there is a rather uncomfortable question sitting underneath all that usage…
Will ordinary consumers actually pay for it?
The latest research from Andreessen Horowitz gives us some fascinating clues.
Only around 4.5% of US consumers in its latest panel had an active paid personal subscription to ChatGPT, Gemini or Claude in August 2026. That number is growing, but it is still a tiny slice of the potential market.
And then things get even more interesting.
The Case For Paying For AI
There is clearly a group of people who are prepared to pay.
In fact, they are prepared to pay a lot.
The top 1% of AI spenders accounted for 19.5% of observed consumer AI spending. These users spent an average of around $903 a month on AI tools.
Who are they?
They tend to be people using AI to build, create, automate and make money.
In other words, they aren’t really buying “AI”.
They’re buying productivity.
That’s the opportunity.
As AI becomes better at actually doing things rather than simply answering questions, the value proposition becomes much easier to understand.
If an AI agent can save me ten hours, create something I can sell, manage a business task or complete a job for me, paying $20, $50 or even $100 suddenly looks very different.
The Case Against Paying For AI
But here’s the problem.
Most people don’t wake up thinking:
“I really need another AI subscription.”
They already have Netflix, Spotify, cloud storage, software subscriptions and a growing pile of other monthly payments.
Why would they add another one?
The research suggests that consumer AI usage is broad, but still relatively shallow. Lots of people experiment with AI.
Far fewer use it every day, and fewer still pay for multiple AI products. Only 13% of people paying for one AI product also pay for another.
That makes the traditional $20-a-month AI subscription model look rather limited.
So What Happens Next?
I think the answer is actually somewhere in the middle.
Consumers will pay for AI, but most won’t necessarily pay for AI itself.
Over the next 6 to 36 months, I expect to see three big changes.
1. AI gets bundled.
Instead of buying “AI”, you’ll get it included with the software, phone, operating system or service you already use.
2. AI gets paid for by the outcome.
Personal agents could move us from paying for a seat to paying for something being done.
Book the holiday.
Find the cheapest flight.
Create the website.
Run the campaign.
Do the paperwork.
3. AI businesses find alternative revenue models.
Advertising, transactions, affiliate revenue, commerce and other forms of indirect monetisation could become much more important.
That’s the really interesting part.
My Bet
I don’t think the future is 100% free AI.
And I don’t think we’re heading towards a world where everyone has ten AI subscriptions.
Instead, I think AI will quietly disappear into the products and services we already pay for.
The winners won’t necessarily be the companies that convince millions of people to buy an AI subscription.
They’ll be the companies that make AI so useful that consumers don’t even think about paying for AI.
They’ll simply pay to get the job done.
And that, I suspect, is where the really big consumer AI market begins.
What do you think?
Would you pay $20 a month for an AI assistant?
Or would you rather pay $20 to have it actually do something for you?
