Every AI Policy Ever · open in the interactive register

Published (non-binding, current)Non-binding · Guidelines / framework

AI Verify – AI Governance Testing Framework and Toolkit

Record SG-02

Issued by
Personal Data Protection Commission (PDPC) and Infocomm Media Development Authority (IMDA)
Jurisdiction
Singapore
Level · region
National · South & Southeast Asia
Published / adopted
2022-05-26
Language
English

Summary

Government AI governance testing framework and software toolkit, launched as a minimum viable product for an international industry pilot. Companies self-test their AI systems with standardised technical tests and process checks and generate reports for stakeholders. The pilot covers 8 of 11 internationally accepted AI ethics principles, grouped in five pillars: transparency on the use of AI, understanding how a model reaches decisions, safety and resilience, fairness, and management and oversight. It sets no ethical standards and does not guarantee that a tested system is free from risks or biases.

Safety topics

evaluation & testingtransparency & labellingaccountability & liability

Status notes

Announced by Minister for Communications and Information Josephine Teo at the World Economic Forum Annual Meeting in Davos. MDDI's newsroom dates the joint press release of the Ministry of Communications and Information, IMDA and the PDPC 26 May 2022; IMDA's copy carries the dateline Davos, 25 May 2022. Developed under the guidance of the Advisory Council on the Ethical Use of AI and Data, it was tested or reviewed by 10 companies before launch. The pilot toolkit worked with a subset of common supervised models (binary classification and regression), mainly tabular datasets and models of about 2GB or less; it did not support unsupervised models.

Sources

Open the official source https://www.mddi.gov.sg/newsroom/sg-launches-first-ai-testing-framework-and-toolkit/

Secondary source: https://www.imda.gov.sg/resources/press-releases-factsheets-and-speeches/press-releases/2022/sg-launches-worlds-first-ai-testing-framework-and-toolkit-to-promote-transparency

Local copy

Archived. Last attempt 2026-10-18 06:48:22 UTC

Official source (third-party copy, not published) archive/SG-02/2026-10-18/official.html
Retrieved 2026-10-18 00:48:29 UTC
SHA-256 97c5e33eb7d644592243dd89ef7108186723add2cd6b9b6ed402eac3391153d0
Linked document (PDF) (third-party copy, not published) archive/SG-02/2026-10-11/document-1.pdf
Retrieved 2026-10-11 14:02:22 UTC
SHA-256 550a609eb68bfa538096a3fe840b64bf8fae9368b630115ab045cc736c4856c8
Linked document (PDF) (third-party copy, not published) archive/SG-02/2026-10-11/document-2.pdf
Retrieved 2026-10-11 14:02:25 UTC
SHA-256 f78f3f90bd7c927cd9509ad02fd8af8280c0d673979b0cc734f8537bf038af24
Secondary source (third-party copy, not published) archive/SG-02/2026-10-18/secondary.html
Retrieved 2026-10-18 00:48:41 UTC
SHA-256 e657fa75b66e31f2cd54022b40036b1bc6fbc9a418555861f35d53e6b8d15c46

Verification

B
Another official source, such as an announcement by the issuer, confirms the title, date and status. checked 2026-10-18 · alternative-official-url-fetched
Recheck of 18 Oct 2026. Stored and read: the joint press release of the Ministry of Communications and Information, IMDA and the PDPC on MDDI's newsroom (copy of 18 Oct 2026; its text is the same as the copy of 11 Oct 2026 except the site's last-updated footer), which gives the name A.I. Verify, the date 26 May 2022, IMDA and PDPC as developers, the MVP and pilot status and the caveats; its Annex B (isomer-user-content.by.gov.sg, Government of Singapore), with the 11 principles, the 8 selected for the MVP and the toolkit's limitations; and IMDA's copy of the press release (secondary source, copy of 18 Oct 2026), whose text layer now includes the body as well as the dateline 'SWITZERLAND, DAVOS – 25 MAY 2022'. All three are announcements about the framework, so the title, date and status rest on another official source, not on the framework's own text (level B). No text of the 2022 testing framework itself was found on an official host; the AI Verify testing framework is now published by the AI Verify Foundation on aiverifyfoundation.sg, which is listed as not official.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-18: 5 field(s) supported.

History

  1. 2026-10-18verification.level: A → B — level computed from the evidence by tools/pipeline.py2026-10-18-refresh-level-rule-1 · Claude Code (claude-opus-5-5)
  2. 2026-10-18notes: Launched as a minimum viable product for an international pilot in May 2022 at t → Announced by Minister for Communications and Information Josephine Teo at the Wo — version 1's account of the AI Verify Foundation and Project Moonshot rests on no stored file of this record and was removed; IMDA's dateline (25 May 2022), the advisory council, the 10 pilot companies and the MVP's limitations added2026-10-18-recheck-south-asia-1 · Claude Code (claude-opus-5-5)
  3. 2026-10-18summary: A government AI governance testing framework and software toolkit. Companies run → Government AI governance testing framework and software toolkit, launched as a m — rewritten from the stored press release and Annex B: version 1's '11 in later versions' rests on no stored file; Annex B gives 11 principles in 5 pillars of which 8 were selected for the MVP2026-10-18-recheck-south-asia-1 · Claude Code (claude-opus-5-5)

Related records

Singapore Model AI Governance Framework Singapore · Published (non-binding, current)

Singapore MGF for Generative AI Singapore · Published (non-binding, current)

Singapore AISI (Digital Trust Centre) Singapore · Published (non-binding, current)

Citation

Personal Data Protection Commission (PDPC) and Infocomm Media Development Authority (IMDA). AI Verify – AI Governance Testing Framework and Toolkit. 2022-05-26. https://www.mddi.gov.sg/newsroom/sg-launches-first-ai-testing-framework-and-toolkit/. Status: Published (non-binding, current); verified 2026-10-18 (level B). Local copy archive/SG-02/2026-10-18/official.html retrieved 2026-10-18 00:48:29 UTC, SHA-256 97c5e33eb7d644592243dd89ef7108186723add2cd6b9b6ed402eac3391153d0. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record SG-02.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.