How people in Czechia use AI at work: AI Pulse Česko 2026

AI Pulse Česko 2026 results from Meetup #2: how often people use AI at work, where it saves time, which barriers hold them back and what happens to the time saved.

At Meetup #2, Barbora Procházková, Zuzana Andělová and Simona Benešová presented the results of AI Pulse Česko 2026. The survey looks at AI through the eyes of people who use it in their everyday work: how often, where it helps, what holds them back and what happens to the time it saves.

← Full Meetup #2 recap

What AI Pulse mapped

We know how many companies are rolling out AI. We know much less about what AI changes for the people themselves. So AI Pulse doesn’t ask companies – it asks people: whose time AI saves, how they really use it and who risks being left behind. The authors describe it as a look beneath the surface of adoption, at the gap between “having AI available” and “using it meaningfully”.

The survey was prepared by the Moderní trh práce (Modern Labour Market) working group of the Inspire & Impact Network ↗ platform. Its authors are Barbora Procházková, Zuzana Andělová, Alena Sochorová, Simona Benešová, Aneta Hochel and Lucie Bášová. It was supported by ČAUI and Women in AI Prague.

How often we use AI

Access to AI no longer sets people apart. Most respondents use it at work regularly.

Source: AI Pulse Česko 2026, Inspire & Impact Network, 2026. Time saved is the respondents’ own estimate in bands.

What’s interesting is what matters less than you might expect: age. The share of people who use AI at least weekly barely differs between age groups.

The authors conclude that differences in AI use have more to do with autonomy and the working environment than with age. Their recommendation for companies and HR: don’t segment training by age, but by roles, conditions and specific barriers.

Where AI really saves time

By their own estimate, respondents together save 1,807 hours a week thanks to AI. The median is 3 hours per person (n = 548). But the savings aren’t spread evenly. They differ most by type of work.

According to the authors, the highest return goes to roles with a self-contained text or technical output and the freedom to redesign their work. The lowest goes to roles bound by process and approvals. A comparison of two related professions shows it well: coaches and mentors use AI daily in 8 of 12 cases, teachers in 2 of 11. One group decides on its own tools, time and budget; the other doesn’t.

The autonomy explanation is the author team’s interpretation, not a proven cause.

What stands in the way of adoption

63% of respondents report at least one barrier. The most common: data concerns (26%), time to learn (26%) and missing rules (15%). The authors stress that barriers are not a trait of people but a diagnosis of the environment.

The strongest number in the whole block shows what happens when barriers add up:

Of the 40 people with three or more barriers, not one saves more than 5 hours a week. And barriers come in pairs. The authors describe the three most common combinations and suggest an intervention for each:

Where change gets stuck

The authors describe an “adoption sandwich”: companies roll out AI through a layer they haven’t given the conditions to.

• CEOs: 2 of 28 lack support; 75% use AI daily.
• Team leads and managers: 16 of 38 lack support; 49% use AI daily; 24% are missing rules.
• Owners: only 4% report missing rules – they set the rules themselves.

Middle management is expected to carry the change into teams but lacks time, mandate, rules and support. The authors see this as the most effective place to intervene – and add explicitly that this doesn’t mean managers are the problem. The survey did not ask about company size.

AI ambassadors work – they just don’t have time

87 of 472 people (18%) described themselves as an ambassador or leader of AI adoption. Compared with the rest of the sample, they use AI daily more often (77% vs 52%) and more of them save over 5 hours a week (31% vs 19%). At the same time, 31% of them say they have no time to learn, and 28 of 87 are worried about company data. Most of them sit at the top: among IT managers, CEOs and HR leadership. In sales, administration, education and PR there are few or none – exactly where adoption is slowest.

Source: AI Pulse Česko 2026, Inspire & Impact Network, 2026. The comparison with the rest of the sample was calculated by the author team from raw data; shares for groups smaller than 20 are indicative.

What happens to the time we save?

If AI really gives us some time back, the next question is what we do with it. Another task? Better work? Learning? Or genuinely freer capacity?

The authors point out that nobody is asking yet. In the open answers they read two things at once: relief – and work intensification. 25 open answers talk about the need to verify outputs and think critically, 10 about overload and higher demands.

“AI saves me time, but it also complicates my work and makes it more demanding.”
AI Pulse Česko 2026 respondent, translated from Czech

The author team offers four directions the freed-up time can take:

Hours are estimates from band midpoints; the authors don’t convert them into jobs or money. The categorisation of open answers, the four directions and “work intensification” are the author team’s interpretation.

And what did the women in the room add?

In smaller groups we talked through four questions the survey authors left us with:

1. Where do you learn about AI?
2. How do you use the time AI saves you?
3. What would you recommend to companies?
4. Does using AI at work also bring you something negative?

In each group someone took notes. This is what the conversations covered.

Where we learn about AI

Most of us learn mainly by trying AI out. Trial, error, another try. Several of us learn about AI with AI itself. Add to that workshops, webinars and courses (Czechitas came up often), newsletters and people we follow on social media.

We also learn a lot from others: from colleagues, in communities, at meetups, in WhatsApp groups, sometimes at home. Company training came up only rarely.

What we miss most is someone to point the way, and examples of how to use AI in our own role.

What we do with the time we save

The most common answer was an honest one: we do more work. We get more tasks done, the to-do list shrinks faster, and new areas appear that we can suddenly handle.

Some of the time goes into learning. There is more room to ask why and how something works, not just hand a task in. And some take part of the time back for themselves: a calm lunch, a coffee, hobbies.

“On the one hand I get more done and my to-do list shrinks faster. On the other, I can go and make myself lunch, a coffee.”
Translated from Czech

What we would recommend to companies

This is where we agreed most: people and sharing. AI ambassadors, a support group of colleagues, peer learning, mentoring. Showing ordinary examples from practice – it doesn’t have to be anything complicated.

Leadership has to stand behind the change and give it time, resources and a mandate. And companies should expect that some people want room to experiment while others need to go step by step.

“It doesn’t have to be anything complicated – just examples of what you can do with AI.”
Translated from Czech

What makes it hard

Almost every group named downsides. The most talked-about were pressure to perform and the fear of missing the train. Overload and scattered attention. Frustration when it doesn’t work and trying things out takes more time than it saves. Responsibility for outputs that aren’t quite ours, and the need to stay on guard with AI all the time.

A quieter feeling came up several times too: that AI is making us lazier about thinking.

What we took away

The conversations in the room matched the AI Pulse numbers quite closely. When nobody decides what happens to saved time, it turns into more work – exactly the “more volume” the authors talked about. And when asked to advise companies, we didn’t ask for more tools but for people: someone to show an example, and time to learn together.

What to take away

The authors closed the block with one sentence: the difference between “I use AI” and “AI really helps me” isn’t made by the tool or by talent. It’s made by conditions – and those are in the company’s hands. As first steps they suggest a mandate for managers, one-page rules, one team process within 30 days, time for ambassadors, and measuring value rather than licences.