Curriculum

Foundations Curriculum (12h)

CASF™ — AI Safety Certified — Foundations Level is the first plain-English AI safety professional certificate. Career-path credential plus a practical outcome: write an AI Safety Policy for a company, school, agency, or deployer. No prerequisites. 12 hours. Verifiable QR + LinkedIn badge.

Format: Video, reading, 10-question quiz each module
Assessment: 60 questions, 60 min, identity-verified, 70% pass, 2 attempts
Frameworks: NIST AI RMF, ISO 42001, EU AI Act, OWASP LLM Top 10

Practical outcome

Certification path: credential + org policy

You do not just pass a quiz. The capstone is an AI Safety Policy leadership can adopt. Teams add EU AI Act deployer policy and org playbooks.

Companies and deployers

Write an AI Safety Policy that maps to NIST AI RMF and EU AI Act deployer duties — not a security checklist dressed up as safety.

Schools and agencies

Produce a one-pager leadership can actually adopt: what you will not deploy, who owns incidents, and how you review models.

The job you already have

Leave with a credential for the job you already have and a policy artifact you can show hiring managers, boards, or a compliance review.

Six modules

M0 · 90 min · Quiz: 5 questions · 80% to pass

Why AI Safety Is Not Security

Free teaser. 90 minutes to finish: 5-min + 25-min + 20-min videos, 30 minutes to write a 1-page policy, and 10 minutes to pass a 5-question quiz (80%). Not the 12-hour path.

Part 1: Safety — Why safety is not security · 5 min
Part 2: Five pillars — Truth, Bias, Privacy, Accountability, Compliance · 25 min
Part 3: Write it live — Fill the 1-page policy template · 20 min

Learning outcomes

Define AI safety vs AI security
Write a 1-page AI Safety Policy
Name the five pillars: Truth, Bias, Privacy, Accountability, Compliance

M1 · 2h · Quiz: 10 questions

How AI Can Fail You

LLM, RAG, and agents in plain English. Seven failure modes. Air Canada and Samsung leaks.

Part 1: LLM, RAG, and agents in plain English · 20 min
Part 2: Seven failure modes · 20 min
Part 3: Air Canada and Samsung leaks · 20 min

Learning outcomes

Explain LLMs, RAG, and agents in plain English
Name seven AI failure modes
Start a risk register from real incidents

M2 · 2h · Quiz: 10 questions

Truth & Bias

Truth and bias controls for deployers who need policy language they can adopt.

Part 1: Truth & Bias · 20 min

Learning outcomes

Spot hallucinations and false authority
Add a Truth section to the policy
Add a Bias section to the policy

M3 · 2h · Quiz: 10 questions

Privacy & Confidentiality

Allowed tools, prohibited data, and confidentiality rules for everyday AI use.

Part 1: Privacy & Confidentiality · 20 min

Learning outcomes

Write an allowed-tools list
Set data rules for prompts
Name what never goes in a prompt: PII, PHI, secrets, client data

M4 · 2h · Quiz: 10 questions

Accountability & Law

Accountability, human review, EU AI Act, and NIST AI RMF without legal jargon.

Part 1: Accountability & Law · 20 min

Learning outcomes

Explain EU AI Act deployer duties in plain English
Apply NIST AI RMF at a beginner level
Assign an owner and review rule for high-risk output

M5 · 2h · Quiz: 10 questions

Ship It: Final Policy

Capstone: turn the one-pager into a 5-page policy leadership can adopt.

Part 1: Ship It: Final Policy · 20 min

Learning outcomes

Assemble the 5-page final policy
Name owner, date, and review cadence
Hand the policy to a manager or HR

Exam domains

Aligned to the same credibility bar as CASO/TAISE — safety-first, not a security dump.

Domain	Weight	Topics
Foundations of AI Safety	20%	Safety vs security, risk taxonomy, lifecycle
Risk Assessment & Threat Modeling	20%	Threat modeling, OWASP LLM Top 10, quantitative and qualitative risk
Safety Integration & Controls	25%	Guardrails, secure coding, monitoring, MLOps controls
Communication & Governance	15%	Writing an org AI Safety Policy, stakeholder engagement, incident reporting
Ethical, Legal & Compliance	20%	Bias and fairness, EU AI Act, ISO 42001, NIST AI RMF, GDPR/CCPA

What you get

12 hours video + reading (lifetime access)
150-question bank, 60-question identity-verified exam, 2 attempts
Verifiable QR + LinkedIn badge at https://www.getaisafetycertified.com/verify/{id}
LinkedIn 1-click add to profile
AI Safety Policy template you can take into a company, school, or agency
2-year validity + $199 renewal (exam again, new GASC ID)

Get free M0: Why AI Safety Is Not Security

Open the free 90-minute module and write your first 1-page AI Safety Policy. Next step: the locked M1–M5 modules, the M6 proctored exam, and the 5-page policy.

Open free M0 Unlock Foundations — $199