Czech elementary school teachers rate their awareness of AI ethics at 3.56 out of 5 — but their ability to actually apply AI tools in the classroom scores just 2.06. That 73% gap is the biggest finding in a 367-teacher survey from the Kralovehradecky Region.
The survey asked teachers across five competency domains — from foundational AI knowledge (K1) to societal impact awareness (K5) — and the results expose a lopsided preparation: teachers have been taught to worry about AI far more than to use it. Ethical awareness (K3) towers over every other domain, while the two most practical skills — application (K2: 2.06) and societal engagement (K5: 2.04) — sit at the bottom.
The chart above makes the imbalance stark. Knowledge (K1: 2.57) and critical evaluation (K4: 2.89) fall in the middle, but the practical bookends — actually deploying tools and understanding their broader implications — are the weakest links. Teachers know AI has risks. They just don't know what to do with it.
Everybody tries ChatGPT. Far fewer stick with it.
Nearly 74% of teachers have tried AI for personal use, and 69% have tried it at work. ChatGPT dominates: 255 teachers (69%) have experimented with it. But only 191 use it regularly — a 25% drop-off. Canva follows a similar pattern, falling from 235 experimenters to 159 regular users. And 80 teachers listed "none" as their regularly used tools, despite having tried several.
This trial-to-habit gap suggests the problem isn't awareness or access — it's competence and confidence. Teachers are experimenting, but without the practical skills to integrate these tools into their workflow, many abandon them.
Using AI at work changes everything
The clearest signal in the data: teachers who actually use AI in their professional practice score dramatically higher across the board. Workplace AI users scored 38% higher on knowledge (K1: 2.80 vs 2.03) and 34% higher on application (K2: 2.23 vs 1.66) than non-users. The gap exists in every competency domain.
This isn't just correlation. It points to a virtuous cycle: hands-on use builds competence, and competence builds the confidence to keep using the tools. The teachers who never cross the threshold from "tried it once" to "use it at work" are left behind — ethical awareness intact but practical ability stalled.
What this means for AI training
Current teacher training programs in the Czech Republic appear to overindex on AI risks and underinvest in practical application. The data suggests that flipping this ratio — giving teachers structured, hands-on tool practice rather than more ethics lectures — could close the competency gap faster. The 56% of teachers who've completed formal AI training still score modestly, which means the training itself may need redesigning, not just expanding. The bottleneck isn't whether teachers care about AI. It's whether they can confidently use it on a Monday morning.