Andriy Onufriyenko/Getty
Artificial intelligence (AI) has reached a tipping point. People have adopted AI at an unprecedented scale, with almost two billion users worldwide, according to an estimate by the US venture capital firm Menlo Ventures.
Also: ChatGPT was downloaded 30 million times last month – but its user base data is more shocking
And yet, very little money is being made, perhaps only $12 billion annually, with most of that figure accounted for by OpenAI.
(Disclosure: Ziff Davis, ZDNET’s parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)
The difference between the enormous number of users and the small amount being earned constitutes “one of the largest and fastest-emerging monetization gaps in recent consumer tech history,” said Menlo Ventures.
Menlo Ventures
The report, “2025: The state of consumer AI,” written by Shawn Carolan, Amy Wu, C.C. Gong, Sam Borja, with unspecified help from Anthropic’s Claude Sonnet large language model, is based on a survey conducted in the US in April for Menlo Ventures by the firm Morning Consult, with the participation of 5,031 US adults.
Also: AI usage is stalling out at work from lack of education and support
The survey found that 61% of US adults had used AI in the prior six months, and “nearly one in five rely on it every day.”
The Menlo Ventures team extrapolated those US-based figures to conclude that as many as 1.8 billion people have used AI tools globally, with as many as 600 million people “engaging daily” with AI. The researchers adjusted global figures to account for worldwide variance in internet access and age distribution, and variance in AI adoption rates by region.
At the same time, the team extrapolated financial figures based on media reports suggesting OpenAI is generating $10 billion in revenue per year, with “approximately 800 million monthly active users” and a ChatGPT Plus subscription price of $20 per month, implying roughly 5% of users pay for premium access.
Overall, only 3% of people are paying to use consumer-grade AI offerings, they noted — “a strikingly low conversion rate.”
Also: Anthropic launches new AI feature to build your own customizable chatbots
The meaning of that disparity between usage and paying for things is that there is a large “white space,” the authors said. That space is an opportunity to build things that people will gravitate toward, and for which they might even pay.
The authors identified “five categories of daily life” in which AI is most used. These categories include “19% of US adults use AI to help write emails,” while “another 18% rely on it to manage to-do lists.”
Menlo Ventures
Those are relatively low rates of adoption, they observed, which suggests to the authors, again, that there is space for new products and services.
“AI is useful everywhere, but there’s still a long way to go to everyday adoption, signaling opportunity for new products to convert casual use before daily habit is locked in,” the report said.
No surprise, there’s a higher percentage for certain activities among those who are regular AI users. For example, those who “routinely engage in a specific activity” tend to carry out the task of “creating images” by 34%.
In fact, the “creative” category of AI use, including image creation, is becoming a must-have for AI users. These creativity-focused tasks also present an opportunity for new products and services.
“The creative category is increasingly AI-native; it’s where consumers most often adopt specialized tools and show a higher willingness to pay,” the report said.
Also: Why ads are coming to your favorite AI bots and you’ve only got yourself to blame
“As it becomes easier to produce polished content, taste — the human eye for what is good, fresh, or interesting — becomes the key differentiator.”
The report continued: “With new image and video models emerging constantly, users care less about which tool they use and focus more on how to control, edit, and refine what AI generates.”
Menlo Ventures
The authors cited some newer tools with AI capabilities, such as Higgsfield and Suno, that “have seen explosive revenue growth as users willingly pay for image generation, audio production, and advanced design.”
At the other end of the spectrum, physical and mental health assistance and management are among the least-used roles for AI. Among US adults, only 14% use AI to research health topics, out of a total of 71% of respondents who generally research health topics.
And out of the 46% for whom tracking nutrition is a regular habit, only 11% use AI for the task.
“[T]hese numbers paint a stark picture: Physical and mental health activities show some of the lowest AI adoption rates among US adults,” the authors wrote. “The bar is high for solutions that handle highly personal information and touch on sensitive topics. In this domain, trust is the real frontier.”
Also: 5 reasons I turn to ChatGPT every day – from faster research to replacing Siri
However, the team concluded that, once again, that gap is an opportunity.
“Companies that combine AI’s data intelligence with trusted human care teams are best positioned to earn patient trust, fit insurance models, and scale high-quality care.”
The survey also included interesting points on the age breakdown of AI use. It revealed that Millennials are the top daily users of AI, more than Gen Z users, which seemed to have surprised the authors.
Menlo Ventures
“While Gen Z (ages 18-28) leads overall AI adoption as expected, Millennials (ages 29-44) emerge as power users, reporting more daily usage — flipping the typical ‘younger = higher usage’ pattern.”
Also, parents turn out to be “the unexpected power users,” which is “one of the most surprising” findings, the authors said.
More than three-quarters (79%) of parents use AI regularly, versus 54% of non-parents. The research suggested parents are “turning to AI for everyday help,” especially Millennial parents, who are “in their prime working years.”
Also: Could this be the reason you cancel your ChatGPT subscription?
The top tasks completed by parents using AI are “learning/understanding new topics” (28%), “taking and organizing notes” (26%), and “managing childcare” (34%).
Overall, usage rises, on average, as parents’ children grow up, which is reflective of increasing complexity in parents’ lives, yet another area of opportunity, the authors concluded.
“The adoption patterns we see among parents show that consumer AI adoption follows life-stage complexity more than just age or income,” they wrote.
“This points to an adoption curve shaped by high-friction ‘life moments’ that savvy companies can target as wedges to build lasting habits.”
