Judgement, demonstrated
Assessed on a deployment the candidate has never seen — scoping it, weighing the applicable requirements, and defending the conclusion. Not recall.
Industry certification body
A competency standard for AI assurance.
The AI Assurance Council certifies the professionals who assess, test, and govern AI systems in regulated and critical industries — from banking supervision to public-sector accountability.
Free · the AIAC AI Literacy Certificate — one assessment, no course, no fee.
Curricula mapped to internationally recognised frameworks
- ISO/IEC 42001AI management systems
- EU AI ActRisk-tiered obligations
- GDPRData protection
- MAS TRM · AIRG · FEATSingapore financial supervision
- DORAEU operational resilience
- IMDA agentic frameworkAgentic AI governance
Curriculum mapping indicates alignment of learning content with published frameworks. It does not constitute endorsement by, or certification under, any standards organisation or regulator.
§ 1 — The credential
What a Council certification requires
Each of these is a published commitment. Where the Council fails one, that is a defect you can raise under the appeals procedure.
- Assessment against a published competency framework
- Examination that tests applied judgement, not recall
- Two-level marking: an automated first pass, a human assessor decides
- Curriculum mapped to the framework or regulation each credential names
- A credential with a unique, publicly verifiable identifier
- Separation of training delivery from examination
- A published code of conduct for credential holders
- Defined validity period with periodic revalidation
- A documented appeals and complaints procedure
§ 2 — Certification tracks
The certification scheme
All track detail →Twelve credentials are published: eleven Fundamentals grades and one endorsement — assurance, governance and compliance credentials anchored to a named framework or regulation, and practice credentials for using AI in a profession — registration opening soon. Every family runs Fundamentals, Professional, and Fellow grades; Professional and Fellow open after each family’s Fundamentals cohort.
FreeAI Literacy CertificateThe awareness tier — one assessment, open to any adult.Open now →Financial servicesAI assurance credentials for regulated financial institutions, where the supervisor is named and the evidence is prescribed.HR & employmentAI recruitment-bias credentials for hiring systems.Public sectorAI assurance credentials for government AI use.LegalAI credentials for the legal profession.Project managementAI credentials for project professionals.
By pillar
§ 3 — The claim
What a credential says about its holder
One standard across functions
Risk, engineering, and audit read a holder’s finding the same way, because every holder was assessed against the same published competency framework.
Checkable by anyone
Every credential has a unique identifier on the public register. An employer verifies it in seconds, without asking the holder or the Council.
Current, not lifetime
A two-year term with revalidation. In a field moving this fast, a credential that cannot expire stops meaning anything.
§ 4 — For organisations
Certify a team against one standard
Organisations adopting AI at scale need a shared vocabulary of risk and a bench of people who can operate assurance processes. The Council works with employers on certifying a team against the published syllabus.
- Cohort enrolment with a single point of administration
- Competency mapping against your existing risk taxonomy
- Progress reporting for L&D and compliance functions
- Start today: the free AI Literacy Certificate, every result on the public register
§ 5 — Insights
Practitioner guides and standards analysis
All insights →Singapore's AI assurance ecosystem, and what it means for practitioners
AI Verify, the Model AI Governance Frameworks, and the world-first agentic AI framework of January 2026 — and what a practitioner is expected to do with them.
NIST AI RMF: a practical walkthrough
What the four functions actually ask you to do, where teams stall, and what changed in 2025–26 — including why there is still no AI RMF 2.0.
How to build an AI risk profile for a deployment
A seven-step method for producing a risk profile a sceptical reader can challenge, with the failure mode at each step and what the finished artefact contains.
Announcements
Follow the scheme as it grows
New credentials, syllabus issues, and examination windows are announced to this list — certification announcements only, no marketing.
- Assessment-based
- Publicly verifiable
- Independently governed
- Framework-mapped