AI Fluency Foundations

What AI is, what it can't do, and your responsibilities. The grounded start of AI work.

Time: ~3 hr
Lessons: 14
Exercises: 14
Level: Foundation

Start Course

14 lessons · enters focused player

What you'll learn

Use the AI vocabulary precisely: differentiate AI from ML from deep learning from generative AI from LLMs from agents.
Describe what AI can and cannot do today, including hallucination, bias, drift, and other failure modes.
Apply the six canonical responsible-AI principles to your own work: fairness, transparency, accountability, privacy, safety, human oversight.
Use the Anthropic 4D framework as a working method: decide what to delegate, describe intent precisely, discern output quality, take diligence on the result.
Make a defensible judgment call on whether to use AI for a given task, with the criteria the canon agrees on.
Name your role and your organization's role under EU AI Act Article 4 and NIST AI RMF GOVERN, and find your employer's AI policy.
Produce a personal AI Fluency Playbook (3 to 5 pages) you keep and reuse: vocabulary glossary, delegation matrix, responsible-AI commitments, risk-and-limits checklist, role-and-policy worksheet.

Description

If you are working with AI in 2026 you need a shared vocabulary, a working sense of what AI can and cannot do, and a clear view of your own responsibilities. This module gives you all three in one place, aligned to the canon every credentialing body has converged on.

Thirteen short lessons plus a vocabulary reference. You start with the stakes (are you in command of the AI around you, or surprised by it?) and a first win running one real task through AI. From there: how the machine actually works and what LLMs can and cannot do, the Anthropic 4 D's, prompt design, the three modes of use (automation, agency, augmentation), discernment and verification, the six responsible-AI principles, what an AI system really is and your role under EU AI Act Article 4 and NIST AI RMF GOVERN, and how to choose AI tools, finishing with a capstone where you assemble your personal AI Fluency Playbook. The vocabulary (AI, machine learning, deep learning, generative AI, LLMs, and agents) sits at the back as a reference to flip to anytime.

Out of scope: model training, cloud architecture depth, MLOps, SDK code, regulatory legal interpretation. Those belong in the cert tracks at /learn, not in literacy. By the end you can pass the AI-literacy section of any major vendor exam blueprint and walk into an Article-4-compliant workplace ready.

Lessons

14 lessons in this micro-course

The deep end

Everything in this course has a technical name — the model families, the training math, the security attack classes. When you're ready for the credential, the professional track teaches all of it.

Explore the professional track