Original AISA research · 91 people · August 2026
Why Don't People Use AI More? [2026 Data]
Why don't people use AI more? AISA asked 91 people what holds them back. See the top barriers and a concrete first step for each one.
Why don't people use AI more? When AISA asked 91 people starting the AI fluency assessment, "What's the biggest thing holding you back from using AI more?" the single largest group (51.6%) said nothing holds them back at all. But the other 48.4% named a real barrier: not knowing where to start, not trusting the output, having no time to learn, or lacking organisational support. This post breaks down each barrier and gives you one concrete first step to move past it.
- Not sure where to start24.2%22
- Don't trust the output8.8%8
- No time to learn12.1%11
- My org doesn't support it3.3%3
- Nothing — I'm all in51.6%47
91 people answered, 6 August to 10 August 2026. Source: AISA research, aisa.to
Why Don't People Use AI More?
The barriers people report are practical, not philosophical. Nearly half of the 91 people AISA asked identified a specific obstacle preventing them from using AI more, even though they were already motivated enough to start an assessment. The answers split across four distinct barriers, each requiring a different response.
What stands out is the gap between intent and action. These are people who showed up to measure their AI fluency. They want to use AI. Something specific is stopping them. Understanding which barrier applies to you is the first step toward clearing it.
The Four Barriers at a Glance
The four barriers people named, in order of prevalence:
- Not sure where to start, 24.2%
- No time to learn, 12.1%
- Don't trust the output, 8.8%
- My org doesn't support it, 3.3%
Each of these is addressable. None of them is "AI isn't useful" or "I don't believe in it." The barriers are about navigation, time, trust, and environment, all things you can work on.
What the "All In" Group Tells Us
The 51.6% who chose "Nothing — I'm all in" represent people who feel no barrier to using AI more. That's worth noting because it means the question isn't whether AI is useful. For the majority, the question is already settled. The interesting work is in the other half, the people who see value but haven't found their way in yet.
Is "Not Sure Where to Start" the Most Common Reason?
Yes. Among people who named a barrier, "not sure where to start" was the most common answer at 24.2%, more than double the next-largest barrier. Nearly one in four people said the problem isn't motivation or trust; it's simply not knowing what to do first.
Why Starting Feels Hard
The AI tool landscape is wide. ChatGPT, Claude, Gemini, Copilot, Perplexity, dozens of specialised tools, the sheer number of options can freeze decision-making. When you don't know which tool to open or what to type into it, the rational response is to do nothing. This is a navigation problem, not a capability problem.
Training gaps make it worse. According to BCG's AI at Work 2025 report (surveying over 10,600 workers across 11 countries), only 36% of employees feel adequately trained in AI use (BCG, June 2025). When most people haven't received structured guidance, "not sure where to start" is the predictable outcome.
What to Do About It
Pick one task you already do repeatedly (summarising meeting notes, drafting a first version of an email, researching a topic) and try it with one AI tool. Don't try to learn "AI" in the abstract. Learn one tool for one task. AISA's free guide on prompting skills walks through how to write your first effective prompt, which is genuinely the smallest useful unit of AI skill.
If you want to know which skill to prioritise, the data from What AI Skill to Learn First? [2026 Data] shows what other people in your position chose.
What If You Do Not Trust the Output?
Distrust of AI output was the barrier for 8.8% of people AISA asked. This is a smaller group than the "not sure where to start" cohort, but it's a qualitatively different problem, these people know how to use AI; they just don't believe what it gives them.
Trust Is a Skill, Not a Feeling
Healthy scepticism toward AI output is actually a strength. AI models produce hallucinations, confident-sounding statements that are factually wrong. The issue isn't that these people are too cautious. It's that they may not yet have a reliable method for checking output, so they default to not using AI at all.
The skill to develop here is verification, not blind trust. Source triangulation (checking an AI claim against independent sources) is the core practice. Once you have a repeatable way to verify output, trust becomes a calibrated judgement rather than a leap of faith.
What to Do About It
Start with low-stakes tasks where you can easily verify the output. Ask AI to summarise a document you've already read. Ask it to draft something you'll review anyway. Each time, note where the output is accurate and where it isn't. You'll build a mental model of when AI is reliable and when it needs checking. AISA's guide on critical thinking with AI covers this verification workflow in detail.
People who struggle with over-reliance face the opposite problem, they trust AI too much. The data in Over-Reliance on AI at Work [2026 Data] shows how common that is. If you're on the sceptical side, you actually have an advantage: you're less likely to accept a hallucination uncritically.

Curious about your AI Fluency?
AISA helps you measure, prove and improve your AI skills — free report in a 20–40 minute chat.
What If You Have No Time to Learn?
"No time to learn" was the barrier for 12.1% of people, the second most common answer after "not sure where to start." This is the time-pressure barrier: people who believe AI could help but can't find space in their schedule to figure it out.
The Training Gap Is Real
The time barrier doesn't exist in a vacuum. Microsoft and LinkedIn's Work Trend Index found that only 39% of people who use AI at work have had AI training from their company (Microsoft, May 2024). When your employer doesn't provide structured training, learning AI becomes a personal project competing with everything else on your plate.
The result is a catch-22: you don't have time to learn AI, so you keep doing tasks manually, which takes more time, which leaves you with even less time to learn AI.
What to Do About It
The fastest way to learn AI is to use it on real work, not to study it in the abstract. Pick a task that takes you 15-30 minutes and try completing it with AI assistance. If it works, you've saved time and learned something. If it doesn't, you've lost 15 minutes and learned what doesn't work, which is also useful.
Task decomposition is the skill that makes this practical. Break a larger task into steps, then identify which step AI could handle. You don't need to automate your whole workflow. You need to find one step where AI saves you five minutes. That five minutes becomes your learning time.
What If Your Organisation Does Not Support AI?
Only 3.3% of people named organisational support as their biggest barrier. It's the smallest group, but it's also the hardest barrier to solve individually because it's environmental rather than personal.
What "Doesn't Support It" Can Mean
Organisational barriers range from explicit ("we've banned AI tools") to implicit ("nobody talks about it, there's no budget, there's no policy"). The BCG data showing only 36% of employees feel adequately trained (BCG, June 2025) suggests that even in organisations that nominally support AI, the support often doesn't reach the individual.
What to Do About It
If your organisation has an explicit ban, respect it, there may be data privacy or compliance reasons. But if the barrier is ambiguity rather than prohibition, you can often start with free tools on non-sensitive tasks. Summarise a public article. Draft a personal email. Use AI for tasks that don't involve company data.
If you're in a position to influence policy, the post on AI Champions Program: Who Teaches AI? [2026] covers how organisations are building internal AI advocacy. Sometimes the barrier isn't that leadership opposes AI, it's that nobody has asked.
What Is the First Step for Each Barrier?
Every barrier has a single, concrete first step. The table below maps each one to an action you can take this week and a resource to support it.
| Barrier | Share | First Step | Resource |
|---|---|---|---|
| Not sure where to start | 24.2% | Pick one repetitive task and try it with one AI tool | Prompting skills guide |
| No time to learn | 12.1% | Use AI on a real 15-minute task instead of studying in the abstract | Using AI at work guide |
| Don't trust the output | 8.8% | Try a low-stakes task where you can verify the result yourself | Critical thinking with AI guide |
| My org doesn't support it | 3.3% | Use free tools on non-sensitive, personal tasks to build skill | Responsible AI use guide |
The Common Thread
Every first step in the table shares one principle: start with a real task, not a course. The people who said "Nothing — I'm all in" didn't necessarily take a class first. They found a use case, tried it, and iterated. The fastest path from stalling to using AI is to do something small and concrete, then do it again.
Measure Where You Stand
If you're not sure which barrier applies to you (or whether you've already moved past it) a structured assessment gives you a baseline. The AISA AI fluency assessment measures 11 criteria across prompting, critical thinking, technical understanding, workflow application, and responsible use. The result is a detailed report showing exactly where your strengths and gaps are, with a score that holds up on retake. It's a 20-minute conversation, not a multiple-choice quiz, and you get a free report and certificate at the end.
Knowing your starting point turns "I should use AI more" into "I need to work on X." That specificity is what makes the difference between stalling and starting.
About this research
This finding comes from AISA's own research. AISA asks a short research question of people as they start the AISA AI fluency assessment, one question at a time. The question in the chart above was asked from 6 August to 10 August 2026, worded exactly as shown and with the answers in the order shown, and 91 people answered it. Each person counts once: when someone answered more than once, their latest answer is the one counted.
More findings from this research are collected in AISA research and data.
How to cite this research: AISA (2026). "Why Don't People Use AI More? [2026 Data]". AISA research. https://aisa.to/blog/why-dont-people-use-ai-more
Related reading: What AI Skill to Learn First? [2026 Data], see which skill other people prioritise when they're just getting started.
Related reading: Over-Reliance on AI at Work [2026 Data], the opposite problem: what happens when people trust AI too much.
Related reading: AI Champions Program: Who Teaches AI? [2026], how organisations build internal AI support from the ground up.
Frequently Asked Questions
Why Do Some People Avoid AI?
The most common reasons are practical, not ideological. Among 91 people AISA asked, the top barriers were not knowing where to start (24.2%), having no time to learn (12.1%), and not trusting AI output (8.8%). These are navigation, time, and trust problems, each with a specific first step to address.
How Do I Start Using AI If I Do Not Know Where to Begin?
Pick one task you already do repeatedly and try it with a single AI tool. Don't attempt to learn everything at once. Summarising notes, drafting emails, or researching a topic are good starting points because you can immediately judge the quality of the output.
How Much Time Does It Take to Learn AI?
You don't need to set aside dedicated study time. The most effective approach is to use AI on a real task that normally takes you 15 to 30 minutes. If the AI helps, you've saved time and learned something simultaneously. Repeat with different tasks and you build fluency through practice rather than coursework.

Curious about your AI Fluency?
AISA helps you measure, prove and improve your AI skills — free report in a 20–40 minute chat.

