AI Literacy Self-Assessment

Use before launching AI initiatives to identify gaps, or after implementation to measure progress. Poll 10–30 people from different groups and roles for an organisational assessment.

Step 1 — Enter your PRESS scores

People-first (P)

Responsible (R)

Explainable (E)

Safe (S)

Sustainable (Su)

Step 2 — Literacy priority framework

Below 4 = crisis (address first) · 4–6 = stable (functional training) · 7+ = strength (advanced learning)

PRESS dimension	If below 4 (crisis)	If 7+ (strength)
People-first	Change management skills — why people fear AI, five resistance types, earning trust	Master collaboration design — teach users, design feedback loops, optimise human-AI teams
Responsible	Regulatory knowledge — protected categories, what bias means, documentation	Higher ethics — fairness beyond rules, advanced bias reduction, ethical leadership
Explainable	Communication skills — metrics to business language, explaining failures plainly	Deep transparency — complex explanation design, audit trail systems
Safe	Data quality knowledge — five problems and red flags	Advanced data mastery — privacy by design, governance, adversarial testing
Sustainable	Cost knowledge — where costs come from, how they explode at scale	Deep technical mastery — advanced serving, distributed systems, cost engineering

Step 3 — Three levels of skill assessment

These three levels build on each other. Most organisations need to start at Level 1.

Level 1: AI literacy — everyone

Everyone understands what AI can and cannot do. This is the base layer.

Can your team explain what AI can and cannot do in plain language?

Do people understand how AI makes decisions (learned patterns, not programmed rules)?

Do people know why AI gives different answers to the same question?

Do people know when to trust AI output and when to check it?

Do people know how to give AI useful feedback that improves the system?

Level 2: AI collaboration — users

Users work effectively with AI tools. Requires Level 1 first.

Can users write clear instructions and prompts that get useful AI outputs?

Do users understand confidence scores and act appropriately on uncertainty?

Do users provide feedback that actually improves AI systems over time?

Are users trained on real-life scenarios with hands-on practice, not just theory?

Level 3: AI development — technical staff

Technical staff can build and maintain AI systems.

Do engineers have foundational AI/ML knowledge?

Have engineers completed hands-on workshops with real data?

Have engineers worked on real projects with guidance?

Is there a clear six-month upskilling path?

Step 4 — Operating model (recommended from your PRESS scores)

Model	When your PRESS scores point here	How it works
Centralised excellence	Sustainable 7+, people-first below 6	Central AI team handles all work.
Distributed empowerment	People-first 7+, sustainable below 6	Each unit has its own AI team.
Hybrid orchestration	Balanced profile — all dimensions 4–6	Small central team + embedded specialists.

Step 5 — Training track (matched to your PRESS priorities)

Track	For PRESS scores	Duration	What it covers
Survival track	Any dimension below 4	2–4 weeks	Focus on the crisis dimension’s specific knowledge gap.
Functional track	All dimensions 4–6	4–8 weeks	Core concepts + role-specific applications.
Advanced track	Any dimension 7+	3–6 months	Deep technical + applied projects + capstone.

Step 6 — The five DNA signs (12–18 months post-implementation)

These signs only become visible 12–18 months after your AI system goes live. For each sign: does this happen naturally, without being prompted?

out of 5 DNA signs present
Signs that happen naturally in your organisation today

Default thinking — "Could AI help?" is the first question for new problems.

Automatic language — AI terms flow naturally across departments.

Embedded processes — Risk assessment happens instinctively.

Natural multiplication — New AI champions emerge without formal assignment.

Invisible integration — AI is "how we work", not "the AI project."

Your literacy summary

Literacy status at a glance

Crisis dimensions

Strength dimensions

Recommended operating model

Recommended training track

Highest skill level achieved

DNA integration score

Your first priority action