AI in Polish companies.
Report 2026
The study “How Polish companies really handle AI”. We ask where companies use AI, who decides about it, whether they measure the effects and whether they are ready for the AI Act.
about 7 minutes · 24 questions · anonymous
Five dimensions of AI maturity.
- 01
Scale of adoption
How often and in how many areas the company uses AI.
- 02
Operational embedding
Whether AI changed processes or was just added to them. Training and employment.
- 03
Measurement and effect
Whether the company measures effects and what impact it sees on operating profit.
- 04
Governance and investment
Who decides about AI and how much the company spends on it.
- 05
Compliance and risk
The AI Act, an AI use policy and control over which tools people use.
Analysis plan
Seven questions the report will answer.
We wrote the hypotheses down before collecting data. We will publish the result of each one, including those that are not confirmed.
- H-01
Do companies that redesigned processes for AI see its impact on results more often than those that added AI to old processes?
- H-02
Does the lack of an AI policy go hand in hand with using tools the company has not approved?
- H-03
Do large companies have more AI pilots but not more processes that actually run on AI?
- H-04
Are companies that measure AI effects more mature regardless of size and industry?
- H-05
What goes together with AI effects more often: how long a company has used AI or how much it spends?
- H-06
Do companies that use AI and those that do not see completely different barriers?
- H-07
Are workforce reductions reported more often by moderately advanced companies than by the most mature ones?
Questionnaire 1.0
Seven blocks, 24 questions.
This is the survey from the inside. Before the AI questions we show a definition of what counts as AI, so everyone answers about the same thing.
- A
About the company
Q-01How many people does your company employ?
4 q. - B
How you use AI
Q-05How often is AI used in your company's work?
5 q. - C
Implementation and people
Q-10What did the company do when introducing AI?
3 q. - D
Effects
Q-13Does the company measure the effects of using AI?
2 q. - E
Who decides and what it costs
Q-15Who in the company is responsible for decisions about AI?
3 q. - F
Compliance and risk
Q-18Do you know which AI Act obligations apply to your company?
3 q. - G
What holds you back and what next
Q-21What limits wider use of AI in your company?
4 q.
A company that does not use AI answers 16 questions. Its answers on barriers and regulatory awareness form a separate chapter of the report, not a discarded group.
AI Maturity Index
One number from 0 to 100, five dimensions.
We calculate a maturity index from the answers. Weights and scoring were written down before data collection, so the result does not depend on what comes out.
Maturity classes
0–20
Observer
21–40
Experimenter
41–60
Adopter
61–80
Operator
81–100
Integrator
Methodology
Rules written down before collecting data.
Definition before measurement
We ask about nothing we have not defined first. The AI definition screen is mandatory.
A base for every number
Every percentage in the report comes with the number of respondents and the full question wording.
Analysis plan before data
Hypotheses and tests were written down before collection started. Anything outside the plan is labelled exploratory.
Honest about the sample
We write about “Polish companies” only for data from a quota-based research panel. Results from the open survey are described as results of the surveyed companies.
Anonymous and separate
We do not collect company names or personal data. Email addresses go to a separate table, never linked to answers.
Open instrument
We publish the full survey with the report, so every chart can be reproduced.
Process
What happens after the survey.
- 1
You fill in the survey
24 questions, about 7 minutes. Anonymous, no company name.
- 2
We close collection and analyse
We follow the plan written down before the start and calculate the AI Maturity Index.
- 3
You get the report
If you leave your email, the report reaches you before public publication.
For participants
The report before publication.
At the end of the survey you can leave your email. Not filling it in? Sign up here and the report arrives on publication day.
Publishers
Who runs the study.
A consultancy for AI implementation in medium and large companies. It starts with order in processes and data, and only then adds AI.
yesfor.ai →
MAI InstituteA research and development institute that combines market data with practitioners’ experience.
mai-inst.com →Questions
Before you start.
How long does it take?
About 7 minutes. If the company does not use AI, the survey is shorter: 16 questions instead of 24.
Who should fill it in?
Ideally someone who knows the decisions about AI tools in the company: the board, directors or a head of department. We ask about your role to know whose perspective it is.
Are the answers anonymous?
Yes. We do not ask for the company name or personal data. The optional company domain is turned into a hash in your browser, and your email is stored separately.
What do I get?
If you leave your email, the report reaches you before publication.
Can I stop and come back?
Yes. Your answers are saved in your browser, so you continue where you left off.
