You have probably heard the story by now. AI is everywhere. Everyone is using it. It is changing everything about how we work, think, and live. Every company is "AI-first." Every job is being transformed. If you are not using AI daily, you are falling behind.
Here is the problem. None of that is true.
Gabriel Weinberg, the founder of DuckDuckGo, just published a piece that pulls together every major dataset on AI adoption in America. His conclusion is simple and backed by hard numbers from Gallup, Microsoft, Datos, and several research institutes.
The reality? About one third of Americans actively use AI. One third use it occasionally. One third never touch it.
That is a very different picture from the one you get from tech conferences, LinkedIn, or the news.
The Data Tells a Different Story
Let me walk through the numbers because they are genuinely surprising.
Microsoft's own telemetry. Microsoft published a study based on anonymized data from actual user behavior across the United States. They defined "using AI" as engaging with major AI services (ChatGPT, Google Gemini, Anthropic Claude, Microsoft Copilot) for at least 90 minutes per month. Their finding: slightly more than 30% of the US working-age population qualifies.
Read that again. By Microsoft's own data, roughly 70% of working-age Americans do not use AI tools for even 90 minutes a month. That is less than 3 minutes a day. And adoption only grew 3 percentage points from the end of 2025 to early 2026.
The Datos study. Datos tracked desktop usage in June 2025. They found that only 21% of desktop computers visited any AI tool more than 10 times in a month. 62% of desktops visited AI tools zero times. Not once.
Searchlight Institute. Their survey found that 58% of Americans have tried or used AI. But that breaks down further. Only 30% use it a few times a month or more. The other 29% have used it once a month or less. "Tried it once" counts as a user in most headlines.
The Argument survey. Their conclusion was blunt: "Most Americans use AI once a week or less."
So when you see headlines claiming AI adoption is skyrocketing, remember what "adoption" means. In many of these surveys, using ChatGPT once to write a wedding toast counts you as an "AI user."
Gen Z Is Not as AI-Obsessed as You Think
Here is where it gets interesting. Gallup tracked Gen Z attitudes toward AI across 2025 and 2026. Gen Z is supposed to be the generation that lives and breathes AI.
The numbers tell a more complicated story:
- 79% to 81% use AI "at least rarely" (so about 1 in 5 Gen Zers never use it at all)
- 31% use it only monthly or every few months
- Anger about AI jumped from 22% to 31% in one year, a 40% relative increase
- Anxiety about AI held steady at around 41-42%
Let me put that anger statistic in perspective. Nearly one in three young people are actively angry about AI. Not indifferent. Not skeptical. Angry. That is a significant shift in just twelve months.
What Are People Actually Worried About?
The Searchlight Institute asked people to name their top concerns about AI. The top three:
- AI will replace jobs and cause unemployment (42%)
- AI will violate people's privacy (35%)
- AI will spread misinformation and lies (33%)
These are not irrational fears cooked up by people who do not understand technology. These are the exact problems we have watched play out in real time. Companies have announced layoffs and cited AI in the same breath. AI tools have been caught scraping personal data. AI-generated misinformation flooded the 2024 election cycle.
People are paying attention. They just do not like what they see.
AI Ranks Below Solar Panels
Here is a number that stuck with me. Researchers asked Americans to rate the net positive societal impact of different technologies. The results:
| Technology | Net Positive Rating |
|---|---|
| Cell phones | +68% |
| The internet | +67% |
| Solar energy | +65% |
| AI | +8% |
| Social media | +7% |
| Crypto | -17% |
AI sits right next to social media. People feel about AI the way they feel about social media. That should worry anyone building AI products.
For context, the internet scores +67%. The same internet that gave us spam, pop-up ads, data breaches, and cyberbullying. People think the internet has been better for society than AI by a margin of 59 percentage points.
The Argument study put it well: "People are not really buying the bullish case for AI that CEOs and boosters alike are selling. The skepticism about AI's effects is real and deep-running. And given how many people use it daily, this is not just an ill-informed set of opinions on something respondents have never seen before."
That last part matters. This is not ignorance. Many of these skeptics have tried AI themselves. They used it. They formed an opinion. The opinion is: meh.
Workers Are Losing Trust the More They Use It
A Fortune report from January 2026, citing ManpowerGroup's Global Talent Barometer, added another layer to this picture. Workers who use AI more frequently actually trust it less over time.
ManpowerGroup's finding: "AI adoption is accelerating, but confidence is collapsing."
Think about that. The people with the most hands-on experience with AI tools are the ones losing faith fastest. That is not how technology adoption usually works. When people buy iPhones, they tend to like them more the longer they use them. When workers start using AI, they start seeing the cracks. The hallucinations. The confident wrong answers. The work that looks good until you actually read it carefully.
This matches what MIT researchers found. They described a widening gap between how much people use AI and how much actual business value it creates. AI tools are not fitting into real workflows as smoothly as the marketing suggests.
The Bubble Problem
Weinberg makes a point that I think is the most important takeaway from all of this. The reason we all think "everyone is using AI" is because we live in a bubble.
If you work in tech, read tech news, follow tech leaders on social media, and attend tech conferences, AI feels unavoidable. Everyone around you is talking about it. Every startup pitch includes it. Every product update mentions it.
But you are not most people. You are maybe 5-10% of the population. The other 90% are living their lives, and AI is not a big part of those lives.
This is a classic echo chamber effect. The same thing happened with crypto in 2021. Inside the crypto bubble, it felt like the whole world was adopting blockchain. Outside the bubble, most people did not care, and eventually the bubble popped.
I am not saying AI is crypto. AI has real, proven utility for specific tasks. But the gap between perception and reality is similar, and that gap has consequences.
Why Most People Are Not Impressed
When you actually talk to people outside the tech bubble, their reasons for not using AI are practical, not ideological.
The tools are often wrong in ways that are hard to catch. If you are not an expert in the subject, you cannot tell when the AI is making things up. And if you are an expert, you spend so much time checking the output that you wonder why you did not just do it yourself.
The tools are also awkward to use for many real tasks. They are great at answering questions, but most jobs are not about answering questions. They are about doing things: filing reports, coordinating with people, managing schedules, fixing problems. AI tools help with the edges of these tasks but not the core.
And then there is the cost. The good AI tools are not cheap. ChatGPT Plus is $20 a month. Claude Pro is $20 a month. For someone making $40,000 a year, that is real money for a tool they will use a few times a month.
What the Numbers Mean for the Future
I want to be clear about something. These numbers do not mean AI is a failure. They mean AI adoption is slower and more uneven than the hype suggests.
We have seen this pattern before. The internet took years to reach mainstream adoption. Smartphones took years. Both technologies eventually became universal, but not on the timeline that early boosters predicted.
AI could follow the same path. The tools will get better. Prices will come down. New use cases will emerge. But right now, in mid-2026, the honest answer is: most people have tried AI, formed a mild opinion, and moved on with their lives.
The companies that succeed with AI will be the ones who understand this. Not the ones who assume everyone is on board. Not the ones who build products for the 5% of power users and ignore everyone else. The ones who figure out what the other 70% actually need and give it to them in a way that does not feel forced.
As Weinberg puts it: "It is not all sunshine and rainbows. It is also clearly not binary, all use or no use, but instead a continuum of AI opinions and use, with a lot of people in the middle."
That middle is where the real market is. And right now, almost nobody is serving it well.
My Take
I use AI tools every day. I find them genuinely useful for certain tasks: drafting, brainstorming, summarizing long documents. But I also catch them being wrong constantly. And I have stopped trusting them for anything that matters without double-checking.
The data in Weinberg's piece matches what I see in my own life. Most of my friends and family outside of tech have tried ChatGPT once or twice. They thought it was neat. Then they went back to doing things the way they always have.
That is not failure. That is the normal adoption curve for any new technology. What is abnormal is the narrative that everyone is already living in an AI-powered future. They are not. And pretending they are will lead to products, policies, and business strategies built on a fantasy.
The truth is less exciting than the hype. But it is a lot more useful.
