EU AI Act Article 4 is in force. Trellis helps you show the AI literacy of your team. How Trellis supports Article 4

Not just who’s using it, but who’s using it well.

AI fluency assessments for teams.

Trellis measures how well your team actually uses AI, not just logins and licences. About 10 minutes for you to set up, about 25 minutes per person, and a team gap report as soon as three people finish.

Ink illustration of a person reaching up to tend a trellis

The problem

Your team has AI tools. Do they use them well?

Most teams have rolled out AI without a way to tell whether it’s helping. Adoption dashboards count usage. Surveys ask people to rate themselves. Neither shows the quality of the work.

Rolled out, not measured

Licences and logins show that people open AI tools. They say nothing about whether the work that comes out is any good.

Training isn’t proof

Finishing a course doesn’t show that someone can judge AI output in their own job, or knows what not to paste into it.

Article 4 now applies

The EU AI Act expects organisations to ensure AI literacy among staff. You need a way to show where your team stands.

What you get

See where your team stands, and what to do next

A team view across five pillars of AI fluency, with a plain-English summary and three next steps you can act on this fortnight. Download it as a PDF for leadership.

Example report with sample data. Managers see team averages only; individuals see their own results.

How it works

From setup to team insight

Privacy by design

Built so people can be honest

People learn more when they aren’t being graded. Trellis reports on the team, not on individuals, and each person decides whether to share their own results.

What you see as a manager

  • Team averages across the five pillars, once three people have finished
  • A plain-English summary of strengths, gaps and next steps
  • A PDF report you can share with leadership
  • Who has completed, so you can follow up
  • Individual results only for people who choose to share them

What stays private

  • Each person’s scores, unless they opt in to share
  • Their quiz answers
  • The AI conversation they paste in
  • Any team result until at least three people have completed

Your data is stored in the EU, in Ireland. The AI conversations people submit are scored by Anthropic’s Claude.

The framework

Five pillars of AI fluency

Grounded in the US Department of Labor AI Literacy Framework and the EU’s DigComp 3.0.

ScopeKnowing what to ask AI to do
CraftKnowing how to ask AI effectively
JudgmentKnowing whether to trust what AI returns
StewardshipKnowing the responsibilities that come with AI use
MomentumKnowing how to keep growing

EU AI Act, Article 4

Article 4 asks organisations to ensure a sufficient level of AI literacy among their staff. Trellis completion records and Stewardship results can support your evidence of this.

Framework alignment

Ready to find out?

About 10 minutes to set up. Share a link with your team. Get your gap report as soon as three people have completed.

Assess your team