Our Mission
We believe AI literacy is the defining professional skill of this decade — and that measuring it honestly is the first step to building it intentionally.
Why a 6-dimension framework?
Most AI assessments treat AI literacy as binary — you either “know AI” or you don't. We disagree. AI skill is multi-dimensional: you can be strong at understanding how AI works (L1) but weak at evaluating its outputs critically (F2). A single score misses everything that matters for workplace readiness.
Our framework separates Literacy (conceptual understanding) from Fluency (practical skill), giving organizations a precise, actionable picture of where to invest.
Built with AI, for AI learning
CPAI is built end-to-end with Anthropic's Claude. Our AI Practice Partner, Adaptive Assessment, Team Gap Synthesis, and Readiness Forecasting all use Claude to provide experiences that would be impossible with static content.
We believe the best way to teach people to work with AI is to show them, not tell them — which is why every learner gets an AI coach that knows their exact scores.
The Assessment Framework
| Code | Dimension | What It Measures | Type |
|---|---|---|---|
| L1 | AI Foundations | Understanding of models, training, capabilities, and limitations | Literacy |
| L2 | Ethics & Critical Thinking | Recognizing bias, hallucinations, privacy risks, and ethical trade-offs | Literacy |
| L3 | Strategic Judgment | Knowing when to use AI, when not to, and how to apply it strategically | Literacy |
| F1 | Prompting Craft | Writing effective prompts that produce consistently high-quality outputs | Fluency |
| F2 | Evaluation & Verification | Critically assessing, fact-checking, and refining AI-generated content | Fluency |
| F3 | Workflow Integration | Designing repeatable processes that embed AI into real work | Fluency |
Built with Anthropic Claude
All AI features in CPAI — from the coaching conversations to the executive ROI reports — are powered by Anthropic's Claude models. We chose Claude for its safety, instruction-following, and long-context reasoning capabilities.