artificial-intelligence.giGlobal AI Monitor

Singapore AI Regulation & Governance

SG11 active modules1 quiet

Last updated · 12 modules · 14 instruments · 54 claims · 52 sources in the cumulative register

Framework status
Model AI Governance Framework (Published 2020-01-21).
Main regulator

RegulatorInfocomm Media Development Authority (IMDA) — AI governance lead

Next milestone
No upcoming dated milestone recorded
Record status
Live · Periodic active record · 52 sources (34 retrieved) · 11 under review · Not yet expert-reviewed
AI-assembled from cited sources
Evidence-gated · How we verify ·
DisclosureNot yet expert-reviewed · Not legal, compliance or investment advice. Facts are machine-assembled from cited public sources. Individual claims have not each been independently verified by a human reviewer. See methodology.

Source tiers: T1 enacted statute, regulation, or binding supervisory order · T2 official regulator guidance, consultation, or registry · T3 international standard-setter or peer-reviewed analysis · T4 practitioner commentary or secondary reporting. Confidence: high = corroborated by retrieved T1/T2 text · medium = single official source or secondary corroboration · low = preliminary or unverified. Full definitions →

Research checked · Updated
Record history

Research as of · Record last changed · Published · Page built

Jump to module (12)

#

Model AI Governance Framework (Published 2020-01-21).

3 instruments (3 current) · 0 obligations · 10 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 247 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Singapore has no single horizontal AI statute. AI-relevant duties are instead distributed across sectoral law and voluntary IMDA/PDPC frameworks, and there is no general, tiered AI-law penalty regime: penalties that touch AI arise only from specific instruments, such as the election-advertising deepfake offence, the Penal Code's AI-generated intimate-image offence, and the Strategic Goods (Control) Act's export-control offences, each carrying its own fixed fine or custodial term. Since July 2026, the Personal Data Protection Commission has required organisations seeking consent to use personal data to train or fine-tune generative AI models to provide AI-specific notifications, an advisory expectation layered onto the Personal Data Protection Act's general consent regime. At the policy level, Singapore's National AI Strategy 2.0 coordinates cross-agency funding, strategic AI initiatives and international policy alignment, operating alongside the AI Verify Testing Framework and Project Moonshot. The centrepiece non-binding governance instrument remains the Model AI Governance Framework, now in its second edition (21 January 2020) and administered jointly by IMDA and the Personal Data Protection Commission as a guide for private-sector organisations on internal governance, human involvement in AI decisions, operations management and stakeholder communication; neither the Framework nor the accompanying AI Verify testing toolkit carries a statutory certification requirement. IMDA and the AI Verify Foundation issued a companion Model AI Governance Framework for Generative AI on 19 June 2024, and followed it with a voluntary Model AI Governance Framework for Agentic AI, first launched 22 January 2026 and updated to version 1.5 by 5 June 2026.

Instruments (3)
  • Model AI Governance Framework Published 2020-01-21
    Non-binding · Official text
  • Model AI Governance Framework for Agentic AI Published 2026-01-22
    Non-binding · Official text
  • Model AI Governance Framework for Generative AI Published 2024-06-19
    Non-binding · Official text
Claims and evidence (10)

  • Singapore's Model AI Governance Framework is a non-binding guide for private-sector organisations on internal governance, human involvement, operations management and stakeholder communication when deploying AI.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • Future of Privacy Forum T3 · Retrieved · retrieved 4 Nov 2026AI Verify, an AI governance testing framework and toolkit, was launched in May 2022.

  • Singapore's National AI Strategy 2.0 coordinates cross-agency funding, strategic AI initiatives and international policy alignment, alongside the AI Verify Testing Framework and Project Moonshot.

    Confidence: Probable · Evidenced
    Evidence and sources

  • Singapore has no single AI statute; scope and roles for AI are defined piecemeal across sectoral law and voluntary IMDA/PDPC frameworks.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • TechDirectory T3 · Retrieved · retrieved 4 Nov 2026Singapore has no single AI law. [verbatim excerpt recorded in source register SRC-SG-008]

  • Singapore has no general, tiered AI-law penalty regime; penalties touching AI arise only from specific instruments such as the Elections Act deepfake offence, the Criminal Law Act's AI-image offence, and the Strategic Goods (Control) Act's export-control offences, each with its own fixed fine or custodial maximum.

    Confidence: Probable (capped: best source is T3) · As of 19 Oct 2026 · Evidenced
    Evidence and sources

  • Singapore's Model AI Governance Framework is a non-binding guide on AI governance for the private sector administered by IMDA/PDPC.

    Confidence: Probable (capped: best source is T3) · Carried forward
    Evidence and sources
    • Future of Privacy Forum T3 · Retrieved · retrieved 4 Nov 2026AI Verify, an AI governance testing framework and toolkit, was launched in May 2022.

  • IMDA and the AI Verify Foundation published the Model AI Governance Framework for Generative AI on 19 June 2024.

    Confidence: Probable (capped: best source is T3) · As of 19 Jun 2024 · Carried forward
    Evidence and sources
    • CASRAI T3 · Retrieved · retrieved 4 Nov 2026IMDA and the AI Verify Foundation published the Model AI Governance Framework for Generative AI on 19 June 2024 — a voluntary, nine-dimension framework that folds Singapore's original 2019/2020 Model AI Governance Framework into its Trusted Development and Deployment dimension.

  • IMDA and the AI Verify Foundation published a voluntary Model AI Governance Framework for Agentic AI, first launched 22 January 2026 and updated to v1.5 by 5 June 2026.

    Confidence: Probable (capped: best source is T1) · As of 22 Jan 2026 · Evidenced
    Evidence and sources
    • CASRAI T3 · Retrieved · retrieved 4 Nov 2026Singapore's IMDA and AI Verify Foundation published a voluntary Model AI Governance Framework for Agentic AI (v1.5, May–June 2026) built around four principles.
    • Infocomm Media Development Authority (IMDA) T1 · Surfaced · retrieved 10 Nov 2026Pre-deployment, test agents for safety and security.

  • From July 2026, the PDPC requires AI-specific notifications where organisations seek consent to use personal data to train or fine-tune generative AI models.

    Confidence: Probable (capped: best source is T3) · As of 1 Jul 2026 · Evidenced
    Evidence and sources
    • Wikipedia T3 · Retrieved · retrieved 4 Nov 2026In July 2026, the PDPC issued advisory guidelines requiring organizations that seek consent to use personal data for training or fine-tuning generative artificial intelligence models to provide AI-specific notifications.

  • As at this cycle, Singapore has no single horizontal AI statute; AI-relevant duties are distributed across sectoral law and voluntary governance frameworks.

    Confidence: Probable (capped: best source is T3) · As of 4 Sep 2026 · Evidenced
    Evidence and sources
    • TechDirectory T3 · Retrieved · retrieved 4 Nov 2026Singapore has no single AI law. [verbatim excerpt recorded in source register SRC-SG-008]

  • Singapore's Model AI Governance Framework (2nd ed., 21 January 2020) and the AI Verify testing toolkit are voluntary and carry no statutory certification requirement.

    Confidence: Probable · As of 21 Jan 2020 · Evidenced
    Evidence and sources
    • Personal Data Protection Commission / IMDA T1 · Surfaced · retrieved 4 Nov 2026The Model Framework provides detailed and readily-implementable guidance to private sector organisations to address key ethical and governance issues when deploying AI solutions.

#

Criminal Law (Miscellaneous Amendments) Act 2025 (In force since 2026-08-17).

4 instruments (4 current) · 2 obligations · 5 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 169 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Singapore's Elections (Integrity of Online Advertising) (Amendment) Act 2024 has been in force since 22 January 2025. Under section 42LA(1), it is an offence to publish, during an election period, online election advertising that digitally depicts a candidate saying or doing something they did not, punishable by a fine of up to S$1,000 and/or imprisonment of up to 12 months, and a social media service that, without reasonable excuse, fails to comply with a corrective direction issued in respect of such an offence is separately liable on conviction to a fine of up to S$1,000,000. Under Singapore's Criminal Law (Miscellaneous Amendments) Act 2025, from 17 August 2026 it is understood to be an offence to produce a non-consensual intimate image, including wholly AI-generated material, punishable by up to two years' imprisonment, a fine, or both. Separately, the Online Safety (Relief and Accountability) Act 2025 creates a statutory tort of 'inauthentic material abuse' for AI-altered or AI-generated material under section 16, but as that section has not yet been commenced.

Instruments (4)
  • Criminal Law (Miscellaneous Amendments) Act 2025 In force since 2026-08-17
    Binding · Official text
  • Elections (Integrity of Online Advertising) (Amendment) Act 2024 In force since 2025-01-22
    Binding · Official text
  • Elections (Integrity of Online Advertising) (Amendment) Act 2024 Valid from 2025-01-22
    Binding
  • Online Safety (Relief and Accountability) Act 2025 Date not recorded
    Legal force not established · Official text

Obligations

  • s.377BE (as amended)

    Prohibition on producing an intimate image, including wholly AI-generated material, of a person without that person's consent.

    Prohibition · Applies to: Other · Criminal Law (Miscellaneous Amendments) Act 2025
  • s.42LA(1)

    Prohibition on publishing, during an election period, online election advertising containing a digitally generated or manipulated (including AI-generated) representation of a candidate saying or doing something they did not.

    Prohibition · Applies to: Other · Elections (Integrity of Online Advertising) (Amendment) Act 2024
Claims and evidence (5)

  • Derived: Probable Singapore's Elections (Integrity of Online Advertising) (Amendment) Act 2024, in force from 22 January 2025, makes it an offence to publish AI-manipulated election advertising depicting a candidate falsely, punishable by a fine up to S$1,000 and/or imprisonment up to 12 months.

    Confidence: Probable · As of 9 Sep 2024 · Evidenced
    Evidence and sources
    • Parliament of Singapore / Attorney-General's Chambers T1 · Surfaced · retrieved 4 Nov 2026Any person who is guilty of an offence under subsection (1) shall be liable on conviction by a District Court to a fine not exceeding $1,000 or to imprisonment for a term not exceeding 12 months or to both.

  • Derived: Probable A social media service that, without reasonable excuse, fails to comply with a corrective direction in respect of an election-deepfake offence is liable on conviction to a fine of up to S$1,000,000.

    Confidence: Probable · As of 9 Sep 2024 · Evidenced
    Evidence and sources
    • Parliament of Singapore / Attorney-General's Chambers T1 · Surfaced · retrieved 4 Nov 2026Any person who is guilty of an offence under subsection (1) shall be liable on conviction by a District Court to a fine not exceeding $1,000 or to imprisonment for a term not exceeding 12 months or to both.

  • Derived: Probable From 17 August 2026, Singapore's Criminal Law (Miscellaneous Amendments) Act 2025 makes it an offence to produce a non-consensual intimate image, including wholly AI-generated material, punishable by up to two years' imprisonment, a fine, or both.

    Confidence: Probable (capped: best source is T3) · As of 14 Aug 2026 · Evidenced
    Evidence and sources
    • Planet Today T3 · Retrieved · retrieved 4 Nov 2026Producing intimate images, including completely AI-generated material, without the depicted person's consent becomes punishable by up to two years' imprisonment, a fine, or both.

  • Derived: Probable Singapore's Online Safety (Relief and Accountability) Act 2025 creates a statutory tort of 'inauthentic material abuse' for AI-altered or AI-generated material (section 16), but as of this cycle that section is not yet commenced.

    Confidence: Probable · As of 25 Nov 2025 · Evidenced
    Evidence and sources

  • Singapore's Elections (Integrity of Online Advertising) (Amendment) Act 2024, in force from 22 January 2025, bans publishing manipulated or AI-generated election advertising depicting a candidate falsely.

    Confidence: Probable · As of 22 Jan 2025 · Evidenced
    Evidence and sources
    • Attorney-General's Chambers, Singapore (Singapore Statutes Online) T1 · Surfaced · retrieved 10 Nov 2026An Act to amend the Parliamentary Elections Act 1954 and the Presidential Elections Act 1991 to prohibit the publication of online election advertising containing certain digitally generated or manipulated content about candidates, and for related purposes.
    • Resemble AI T3 · Retrieved · retrieved 10 Nov 2026Up to S$1,000 fine and 12 months’ imprisonment for individuals
    • Elections Department Singapore T1 · Retrieved · retrieved 10 Nov 2026The Elections (Integrity of Online Advertising) (Amendment) Act 2024 comes into operation on 22 January 2025.

#

Model AI Governance Framework for Agentic AI (Version 1.5) (Valid from 2026-05-20).

1 instrument (1 current) · 0 obligations · 1 claim · 0 milestones

Standing sub-brief · as of 12 Nov 2026 50 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Singapore has no binding high-risk-system classification regime. The only standing instrument addressing high-stakes AI deployment is IMDA's Model AI Governance Framework for Agentic AI (version 1.5, effective from 20 May 2026), which recommends human approval for high-stakes autonomous agent actions as voluntary, non-binding guidance rather than an enforceable high-risk tier.

Instruments (1)
  • Model AI Governance Framework for Agentic AI (Version 1.5) Valid from 2026-05-20
    Non-binding
Claims and evidence (1)

  • IMDA's Model AI Governance Framework for Agentic AI recommends human approval for high-stakes autonomous agent actions, as voluntary non-binding guidance.

    Confidence: Probable · As of 20 May 2026 · Evidenced
    Evidence and sources

#

IMDA and the AI Verify Foundation published a voluntary nine-dimension Model AI Governance Framework for Generative AI on 19 June 2024, covering data quality, transparency, incident reporting, security, content provenance, testing and assurance.

0 instruments (0 current) · 0 obligations · 3 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 131 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Singapore's Strategic Goods (Control) Act 2002 has been in force since 1 January 2003 and requires permits for the export, transhipment, transit, intangible transfer and brokering of listed strategic goods and technology, a category that includes advanced AI processors meeting listed technical specifications; a company has been fined for breach of these permit requirements. At the model-governance level, IMDA and the AI Verify Foundation published a voluntary nine-dimension Model AI Governance Framework for Generative AI on 19 June 2024, covering data quality, transparency, incident reporting, security, content provenance, testing and assurance, and followed it with a voluntary Model AI Governance Framework for Agentic AI, launched 22 January 2026 and updated to version 1.5 in May-June 2026, built around four principles for bounding risk and ensuring human accountability in agentic AI deployment.

Claims and evidence (3)

  • IMDA and the AI Verify Foundation published a voluntary nine-dimension Model AI Governance Framework for Generative AI on 19 June 2024, covering data quality, transparency, incident reporting, security, content provenance, testing and assurance.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • CASRAI T3 · Retrieved · retrieved 4 Nov 2026IMDA and the AI Verify Foundation published the Model AI Governance Framework for Generative AI on 19 June 2024 — a voluntary, nine-dimension framework that folds Singapore's original 2019/2020 Model AI Governance Framework into its Trusted Development and Deployment dimension.

  • IMDA's Model AI Governance Framework for Agentic AI, launched 22 January 2026 and updated to v1.5 in May-June 2026, is a voluntary framework built around four principles for bounding risk and ensuring human accountability in agentic AI deployment.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • CASRAI T3 · Retrieved · retrieved 4 Nov 2026Singapore's IMDA and AI Verify Foundation published a voluntary Model AI Governance Framework for Agentic AI (v1.5, May–June 2026) built around four principles.

  • Derived: Probable The Strategic Goods (Control) Act 2002, in force since 1 January 2003, requires permits for the export, transhipment, transit, intangible transfer and brokering of listed strategic goods and technology, including advanced AI processors meeting listed technical specifications; a company was fined over S$1.1 million and individuals S$35,000 and S$45,000 for unpermitted export.

    Confidence: Probable · As of 24 Apr 2025 · Evidenced
    Evidence and sources
    • Steptoe Trade Compliance Resource Hub T2 · Retrieved · retrieved 4 Nov 2026The SGCA regulates the transfer of strategic goods and technologies, such as dual-use goods that are capable of military application.
    • Ministry of Trade and Industry T1 · Surfaced · retrieved 4 Nov 2026Within Singapore, the transfer and brokering of strategic goods and technology is governed by the Strategic Goods (Control) Act.

#

Criminal Law (Miscellaneous Amendments) Act 2025 (Valid from 2025-12-30).

6 instruments (6 current) · 0 obligations · 12 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 322 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Singapore's Elections (Integrity of Online Advertising) (Amendment) Act 2024 has been in force since 22 January 2025. Section 42LA(1) makes it an offence to publish, during an election period, online election advertising that digitally depicts a candidate saying or doing something they did not, punishable by a fine of up to S$1,000 and/or imprisonment of up to 12 months, and a social media service that, without reasonable excuse, fails to comply with a Returning Officer's corrective direction under section 42LA is separately liable on conviction to a fine of up to S$1,000,000. The Criminal Law (Miscellaneous Amendments) Act 2025 extends the Penal Code's definition of 'intimate image' to wholly AI- or synthetically-generated depictions, with its provisions commencing in stages on 30 December 2025 and 17 August 2026; from the later date it is an offence to produce such an image of a person without consent, punishable by up to two years' imprisonment, a fine, or both. The Online Safety (Relief and Accountability) Act 2025 established the Online Safety Commission and, on 29 June 2026, is reported to have brought into force a first phase of statutory online-harm torts, including an 'inauthentic material abuse' category covering AI-generated content, alongside a related section 86 civil tort for communicating AI-generated or otherwise inauthentic material likely to cause harassment. Reporting on commencement is not fully consistent: other material in the record states that the inauthentic-material-abuse provision at section 16 had, as, not yet been commenced, and separate commentary dated 27 August 2026 likewise describes that provision — subject to a labelling carve-out — as still uncommenced; both positions are recorded here pending resolution. In parallel, IMDA and the AI Verify Foundation published the Model AI Governance Framework for Generative AI on 19 June 2024, and in July 2026 the Personal Data Protection Commission issued advisory guidance requiring organisations that seek consent to use personal data for training or fine-tuning generative AI models to provide AI-specific notifications.

Instruments (6)
  • Criminal Law (Miscellaneous Amendments) Act 2025 Valid from 2025-12-30
    Binding
  • Elections (Integrity of Online Advertising) (Amendment) Act 2024 In force since 2025-01-22
    Binding · Official text
  • Elections (Integrity of Online Advertising) (Amendment) Act 2024 Valid from 2025-01-22
    Binding
  • Model AI Governance Framework for Generative AI Published 2024-06-19
    Non-binding · Official text
  • Online Safety (Relief and Accountability) Act 2025 Date not recorded
    Legal force not established · Official text
  • Online Safety (Relief and Accountability) Act 2025 Valid from 2026-06-29
    Binding
Claims and evidence (12)

  • In July 2026 the PDPC issued advisory guidelines requiring organisations seeking consent to use personal data for training or fine-tuning generative AI models to provide AI-specific notifications.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • Wikipedia T3 · Retrieved · retrieved 4 Nov 2026In July 2026, the PDPC issued advisory guidelines requiring organizations that seek consent to use personal data for training or fine-tuning generative artificial intelligence models to provide AI-specific notifications.

  • Derived: Probable It is an offence under s.42LA(1) to publish, during an election period, online election advertising that digitally depicts a candidate saying or doing something they did not, with a fine of up to S$1,000 and/or up to 12 months' imprisonment, in force from 22 January 2025.

    Confidence: Probable · As of 9 Sep 2024 · Evidenced
    Evidence and sources
    • Elections Department Singapore T1 · Retrieved · retrieved 10 Nov 2026The Elections (Integrity of Online Advertising) (Amendment) Act 2024 comes into operation on 22 January 2025.
    • Parliament of Singapore / Attorney-General's Chambers T1 · Surfaced · retrieved 4 Nov 2026Any person who is guilty of an offence under subsection (1) shall be liable on conviction by a District Court to a fine not exceeding $1,000 or to imprisonment for a term not exceeding 12 months or to both.
    • Attorney-General's Chambers, Singapore (Singapore Statutes Online) T1 · Surfaced · retrieved 10 Nov 2026An Act to amend the Parliamentary Elections Act 1954 and the Presidential Elections Act 1991 to prohibit the publication of online election advertising containing certain digitally generated or manipulated content about candidates, and for related purposes.
    • Resemble AI T3 · Retrieved · retrieved 10 Nov 2026Up to S$1,000 fine and 12 months’ imprisonment for individuals

  • Derived: Probable A social media service that, without reasonable excuse, fails to comply with a Returning Officer's corrective direction under s.42LA is liable on conviction to a fine of up to S$1,000,000.

    Confidence: Probable (capped: best source is T3) · As of 9 Sep 2024 · Evidenced
    Evidence and sources
    • Planet Today T3 · Retrieved · retrieved 4 Nov 2026Producing intimate images, including completely AI-generated material, without the depicted person's consent becomes punishable by up to two years' imprisonment, a fine, or both.

  • Derived: Probable From 17 August 2026, the Penal Code's definition of 'intimate image' is extended to wholly AI/synthetically-generated depictions, and it is a new offence to produce such an image of a person without consent, punishable by up to two years' imprisonment, a fine, or both.

    Confidence: Probable (capped: best source is T3) · As of 14 Aug 2026 · Evidenced
    Evidence and sources
    • Dentons Rodyk T3 · Surfaced · retrieved 4 Nov 2026the Act is currently in force for some of its provisions, and section 16 dealing with this form of online harm is not yet in force.
    • Reed Smith T3 · Retrieved · retrieved 4 Nov 2026Under section 243 of the Copyright Act, “computational data analysis,” in relation to a work or a recording of a protected performance, includes

  • Derived: Probable OSRA creates a statutory tort and harm category of 'inauthentic material abuse' (s.16/s.86) covering AI-altered or AI-generated material likely to cause harassment, alarm, distress or humiliation, subject to a labelling carve-out; as at 27 August 2026 secondary commentary states this provision had not yet been brought into force even though other parts of the Act are operational.

    Confidence: Disputed (capped: best source is T3) · As of 27 Aug 2026 · Evidenced
    Evidence and sources
    • Lexology T3 · Surfaced · retrieved 4 Nov 2026Singapore's TDM exception has stayed unchanged since 2021, allowing AI training on copyrighted works without an opt-out.
    • Infinity Legal LLC T3 · Retrieved · retrieved 10 Nov 2026This exception only applies to acts performed in Singapore.

  • IMDA and the AI Verify Foundation published the Model AI Governance Framework for Generative AI on 19 June 2024.

    Confidence: Probable (capped: best source is T3) · As of 19 Jun 2024 · Carried forward
    Evidence and sources
    • CASRAI T3 · Retrieved · retrieved 4 Nov 2026IMDA and the AI Verify Foundation published the Model AI Governance Framework for Generative AI on 19 June 2024 — a voluntary, nine-dimension framework that folds Singapore's original 2019/2020 Model AI Governance Framework into its Trusted Development and Deployment dimension.

  • Derived: Probable Singapore's Elections (Integrity of Online Advertising) (Amendment) Act 2024, in force from 22 January 2025, makes it an offence to publish AI-manipulated election advertising depicting a candidate falsely, punishable by a fine up to S$1,000 and/or imprisonment up to 12 months.

    Confidence: Probable · As of 9 Sep 2024 · Evidenced
    Evidence and sources
    • Parliament of Singapore / Attorney-General's Chambers T1 · Surfaced · retrieved 4 Nov 2026Any person who is guilty of an offence under subsection (1) shall be liable on conviction by a District Court to a fine not exceeding $1,000 or to imprisonment for a term not exceeding 12 months or to both.

  • Derived: Probable Singapore's Online Safety (Relief and Accountability) Act 2025 creates a statutory tort of 'inauthentic material abuse' for AI-altered or AI-generated material (section 16), but as of this cycle that section is not yet commenced.

    Confidence: Probable · As of 25 Nov 2025 · Evidenced
    Evidence and sources

  • Singapore's Elections (Integrity of Online Advertising) (Amendment) Act 2024, in force from 22 January 2025, bans publishing manipulated or AI-generated election advertising depicting a candidate falsely.

    Confidence: Probable · As of 22 Jan 2025 · Evidenced
    Evidence and sources
    • Attorney-General's Chambers, Singapore (Singapore Statutes Online) T1 · Surfaced · retrieved 10 Nov 2026An Act to amend the Parliamentary Elections Act 1954 and the Presidential Elections Act 1991 to prohibit the publication of online election advertising containing certain digitally generated or manipulated content about candidates, and for related purposes.
    • Resemble AI T3 · Retrieved · retrieved 10 Nov 2026Up to S$1,000 fine and 12 months’ imprisonment for individuals
    • Elections Department Singapore T1 · Retrieved · retrieved 10 Nov 2026The Elections (Integrity of Online Advertising) (Amendment) Act 2024 comes into operation on 22 January 2025.

  • OSRAA established the Online Safety Commission and brought the first phase of statutory online-harm torts, including inauthentic material abuse covering AI-generated content, into force on 29 June 2026.

    Confidence: Probable · As of 29 Jun 2026 · Evidenced
    Evidence and sources

  • OSRAA section 86 creates a civil tort for communicating AI-generated or otherwise inauthentic material likely to cause harassment.

    Confidence: Probable · As of 16 Oct 2026 · Evidenced
    Evidence and sources
    • Singapore Statutes Online (Attorney-General's Chambers) T1 · Surfaced · retrieved 10 Nov 202685 Tort of online impersonation 86 Tort of inauthentic material abuse 87 Tort of online instigation of disproportionate harm
    • Covenant Chambers T3 · Surfaced · retrieved 10 Nov 2026Inauthentic Material Abuse (Section 86): Requires proving communication of material known/reasonably believed to be inauthentic, likely causing harassment.

  • The Criminal Law (Miscellaneous Amendments) Act 2025 criminalises producing or possessing non-consensual intimate images, including wholly AI-generated ones, with provisions commencing 30 December 2025 and 17 August 2026.

    Confidence: Probable · As of 17 Aug 2026 · Evidenced
    Evidence and sources
    • Ministry of Home Affairs (MHA), Singapore T1 · Retrieved · retrieved 10 Nov 2026The Criminal Law (Miscellaneous Amendments) Act 2025 ("the Act") was passed by Parliament on 4 November 2025.
    • AsiaOne T2 · Retrieved · retrieved 10 Nov 2026the definition of intimate images will now include materials that are generated to depict a person without altering an original image or recording.
    • Attorney-General's Chambers, Singapore (Singapore Statutes Online) T1 · Surfaced · retrieved 10 Nov 2026Sections 15, 16, 17, 24 to 28, 40 to 62, 65 to 71, 75 and 76 of the Criminal Law (Miscellaneous Amendments) Act 2025 come into operation on 17 August 2026.
    • The Sun (Malaysia) T2 · Retrieved · retrieved 10 Nov 2026the offence is punishable with mandatory imprisonment of up to two years, and the offender shall also be liable to a fine or caning

#

Copyright Act 2021 (In force since 2021-11-21).

2 instruments (2 current) · 1 obligation · 3 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 77 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Singapore's Copyright Act 2021 has been in force since 21 November 2021. Its computational data analysis exception under section 244 permits copying of a work for computational data analysis, including AI model training, without the rightsholder's consent, subject to five conditions including lawful access to the source material and a restriction against supplying the copy to others except for result-verification or collaborative purposes; the exception carries no opt-out mechanism for rightsholders and is subject to territorial limits.

Instruments (2)
  • Copyright Act 2021 In force since 2021-11-21
    Binding · Official text
  • Copyright Act 2021 Valid from 2021-11-21
    Binding

Obligations

  • s.244(2)

    A person performing computational data analysis (including AI training) on copyrighted works must have lawful access to the source material and must not supply the copy to others except for result-verification or collaborative research relating to the same purpose.

    Other · Applies to: Other · Copyright Act 2021
Claims and evidence (3)

  • Singapore's Computational Data Analysis exception (in force since 21 November 2021) permits copying of a work for computational data analysis, including AI training, subject to five conditions including lawful access to the source material, with no opt-out mechanism for rightsholders and a territorial limit to acts performed in Singapore.

    Confidence: Probable · As of 30 Sep 2026 · Evidenced
    Evidence and sources

  • Singapore's Copyright Act 2021 computational-data-analysis exception (in force since 21 November 2021) permits AI training on copyrighted works without an opt-out mechanism, subject to conditions and territorial limits.

    Confidence: Probable (capped: best source is T3) · As of 30 Sep 2026 · Evidenced
    Evidence and sources
    • Reed Smith T3 · Retrieved · retrieved 4 Nov 2026Under section 243 of the Copyright Act, “computational data analysis,” in relation to a work or a recording of a protected performance, includes
    • Lexology T3 · Surfaced · retrieved 4 Nov 2026Singapore's TDM exception has stayed unchanged since 2021, allowing AI training on copyrighted works without an opt-out.
    • Infinity Legal LLC T3 · Retrieved · retrieved 10 Nov 2026This exception only applies to acts performed in Singapore.

  • Singapore's Copyright Act 2021 s.244 permits computational data analysis, including AI model training, on copyright works without consent, subject to lawful access.

    Confidence: Probable (capped: best source is T3) · As of 21 Nov 2021 · Evidenced
    Evidence and sources
    • Singapore Academy of Law T3 · Surfaced · retrieved 10 Nov 2026On 21 November 2021, the Copyright Act 2021 (Act) came into force in Singapore.
    • Infinity Legal LLC T3 · Retrieved · retrieved 10 Nov 2026This exception only applies to acts performed in Singapore.

#

Model AI Governance Framework for Agentic AI (Version 1.5) (Valid from 2026-05-20).

3 instruments (3 current) · 1 obligation · 14 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 330 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Singapore's Strategic Goods (Control) Act 2002 requires a permit, under section 4A, for the export, transhipment, transit, intangible transfer and brokering of listed strategic goods and technology, a category that includes advanced AI processors meeting listed technical specifications. In the financial sector, the Monetary Authority of Singapore's Fairness, Ethics, Accountability and Transparency Principles, issued in 2018, guide financial institutions' use of AI and data analytics, with the MAS-led Veritas consortium providing an optional assessment methodology that firms may use and adapt against the Principles; following a 2024 thematic review of banks, MAS published an Information Paper on 5 December 2024 setting out good-practice supervisory expectations for AI and generative-AI model risk management, covering governance, AI inventories and materiality assessment, and development, validation, deployment and monitoring. The Personal Data Protection Commission published Advisory Guidelines on the use of personal data in AI recommendation and decision systems on 1 March 2024, clarifying how Personal Data Protection Act consent and notification obligations apply to organisations that use personal data to develop, test, monitor and deploy such systems in business-to-consumer contexts, while expressly excluding generative AI training and deployment use cases. There is no legislation governing the use of AI in Singapore's public sector: GovTech and the Smart Nation and Digital Government Office instead issue a Public Sector AI Playbook and a Responsible AI Playbook for whole-of-government AI system-integration projects, both administrative rather than statutory instruments. IMDA's voluntary Model AI Governance Framework for Agentic AI, launched 22 January 2026 and updated to version 1.5 in May-June 2026, is built around four principles for bounding risk and ensuring human accountability in agentic AI deployment, and recommends human approval for high-stakes autonomous agent actions. Singapore's National AI Strategy 2.0, launched 4 December 2023, is a non-binding whole-of-government policy strategy committing the Government to accelerate public-sector AI adoption and run sector-specific AI programmes under the Smart Nation Group together with IMDA, AI Singapore, A*STAR and sector regulators, with priorities refreshed in an announcement on 20 May 2026.

Instruments (3)
  • Model AI Governance Framework for Agentic AI (Version 1.5) Valid from 2026-05-20
    Non-binding
  • National AI Strategy 2.0 Published 2023-12-04
    Non-binding · Official text
  • Strategic Goods (Control) Act 2002 In force since 2003-01-01
    Binding · Official text

Obligations

  • s.4A

    A permit is required before export, transhipment, transit, intangible transfer of technology, or brokering of goods or technology listed as strategic goods or strategic goods technology, including advanced AI processors meeting the listed technical specifications.

    Authorisation · Applies to: Other · Strategic Goods (Control) Act 2002
Claims and evidence (14)

  • IMDA's Model AI Governance Framework for Agentic AI, launched 22 January 2026 and updated to v1.5 in May-June 2026, is a voluntary framework built around four principles for bounding risk and ensuring human accountability in agentic AI deployment.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • CASRAI T3 · Retrieved · retrieved 4 Nov 2026Singapore's IMDA and AI Verify Foundation published a voluntary Model AI Governance Framework for Agentic AI (v1.5, May–June 2026) built around four principles.

  • MAS's FEAT Principles (2018) guide financial institutions' use of AI/data-analytics, and the MAS-led Veritas consortium provides an optional assessment methodology that firms may use and adapt against FEAT.

    Confidence: Confirmed · As of 4 Feb 2022 · Evidenced
    Evidence and sources
    • Monetary Authority of Singapore T1 · Retrieved · retrieved 4 Nov 2026Veritas aims to enable financial institutions to evaluate their AIDA-driven solutions against the principles of fairness, ethics, accountability and transparency (FEAT) that MAS co-created with the financial industry in late 2018
    • Monetary Authority of Singapore T1 · Retrieved · retrieved 4 Nov 2026This information paper sets out good practices for AI and Generative AI model risk management that were observed during the review, focusing on those relating to governance and oversight, key risk management systems and processes, and development and deployment of AI.

  • Following a 2024 thematic review of banks, MAS published good-practice expectations (5 December 2024) for AI and Generative AI model risk management, covering governance, AI inventories and materiality assessment, and development/validation/deployment/monitoring.

    Confidence: Probable · As of 5 Dec 2024 · Evidenced
    Evidence and sources
    • Parliament of Singapore / Attorney-General's Chambers T1 · Surfaced · retrieved 4 Nov 2026Any person who is guilty of an offence under subsection (1) shall be liable on conviction by a District Court to a fine not exceeding $1,000 or to imprisonment for a term not exceeding 12 months or to both.

  • PDPC's 1 March 2024 Advisory Guidelines clarify how PDPA consent and notification obligations apply to organisations using personal data to develop, test, monitor and deploy AI recommendation/decision systems in B2C contexts, expressly excluding generative AI training/deployment.

    Confidence: Probable · As of 4 Mar 2024 · Evidenced
    Evidence and sources
    • Personal Data Protection Commission T1 · Surfaced · retrieved 4 Nov 2026The PDPC has published the Advisory Guidelines on use of Personal Data in AI recommendation and Decision Systems.

  • National AI Strategy 2.0, launched 4 December 2023, commits the Singapore Government to accelerate public-sector AI adoption and run sector-specific AI programmes under the Smart Nation Group with IMDA, AI Singapore, A*STAR and sector regulators.

    Confidence: Probable (Capped from Confirmed: best source is T3) · As of 4 Dec 2023 · Evidenced
    Evidence and sources

  • There is no legislation or hard law regulating the use of AI in Singapore's public sector; GovTech/SNDGO instead issue a Public Sector AI Playbook and a Responsible AI Playbook for Whole-of-Government AI system-integration projects, both administrative rather than statutory.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • AI Law Guide T3 · Retrieved · retrieved 4 Nov 2026It co-issues the Model AI Governance Frameworks (with PDPC and AI Verify Foundation), runs the Generative AI Evaluation Sandbox, drives the National Multimodal LLM Programme

  • MAS's FEAT Principles (2018) and the Veritas Initiative provide a voluntary assessment methodology for financial institutions' use of AI/data analytics.

    Confidence: Confirmed · As of 12 Nov 2018 · Evidenced
    Evidence and sources
    • Monetary Authority of Singapore T1 · Retrieved · retrieved 4 Nov 2026Veritas aims to enable financial institutions to evaluate their AIDA-driven solutions against the principles of fairness, ethics, accountability and transparency (FEAT) that MAS co-created with the financial industry in late 2018

  • MAS published an Information Paper (5 December 2024) setting out supervisory-expectation good practices for AI and generative-AI model risk management at banks and financial institutions.

    Confidence: Confirmed · As of 5 Dec 2024 · Evidenced
    Evidence and sources
    • Monetary Authority of Singapore T1 · Retrieved · retrieved 4 Nov 2026This information paper sets out good practices for AI and Generative AI model risk management that were observed during the review, focusing on those relating to governance and oversight, key risk management systems and processes, and development and deployment of AI.

  • On 1 March 2024 the PDPC published Advisory Guidelines on the use of personal data in AI recommendation and decision systems, applicable to organisations using personal data to develop and deploy such systems.

    Confidence: Probable (capped: best source is T1) · As of 1 Mar 2024 · Evidenced
    Evidence and sources
    • Personal Data Protection Commission T1 · Surfaced · retrieved 4 Nov 2026The PDPC has published the Advisory Guidelines on use of Personal Data in AI recommendation and Decision Systems.
    • Bird & Bird T3 · Retrieved · retrieved 4 Nov 2026The Advisory Guidelines apply to any organisation which collects and/or uses personal data to develop and deploy AI Systems.

  • Singapore launched its National AI Strategy 2.0 on 4 December 2023, a whole-of-government policy strategy for public-sector AI adoption.

    Confidence: Confirmed · As of 4 Dec 2023 · Evidenced
    Evidence and sources

  • Singapore's public-sector use of AI is governed by administrative playbooks (GovTech/SNDGO), not by legislation.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • Mondaq T3 · Surfaced · retrieved 4 Nov 2026There is no legislation or hard law regulating the use of AI in the public sector.

  • Derived: Probable Singapore's Strategic Goods (Control) Act 2002 requires a permit for export, transhipment, transit and brokering of listed strategic goods and technology, including advanced AI processors meeting listed technical specifications.

    Confidence: Probable · As of 24 Apr 2025 · Evidenced
    Evidence and sources
    • Steptoe Trade Compliance Resource Hub T2 · Retrieved · retrieved 4 Nov 2026The SGCA regulates the transfer of strategic goods and technologies, such as dual-use goods that are capable of military application.

  • IMDA's Model AI Governance Framework for Agentic AI recommends human approval for high-stakes autonomous agent actions, as voluntary non-binding guidance.

    Confidence: Probable · As of 20 May 2026 · Evidenced
    Evidence and sources

  • Singapore's National AI Strategy 2.0 is a non-binding policy framework refreshed with updated priorities announced 20 May 2026.

    Confidence: Probable · As of 20 May 2026 · Evidenced
    Evidence and sources

#

The Cyber Security Agency of Singapore released an addendum on securing agentic AI systems for public consultation on 17 June 2026.

0 instruments (0 current) · 0 obligations · 4 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 55 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

The Cyber Security Agency of Singapore's Guidelines and Companion Guide on Securing AI Systems, launched 15 October 2024, recommend a voluntary lifecycle approach to AI security covering development, deployment and operations. On 17 June 2026, the Cyber Security Agency released, for public consultation, an addendum extending these 2024 AI security guidelines to agentic AI systems.

Claims and evidence (4)

  • The Cyber Security Agency of Singapore released an addendum on securing agentic AI systems for public consultation on 17 June 2026.

    Confidence: Confirmed · Evidenced
    Evidence and sources
    • Cyber Security Agency of Singapore T1 · Retrieved · retrieved 4 Nov 2026CSA releases an addendum to support system owners in securing agentic AI systems for public consultation.

  • CSA's Guidelines and Companion Guide on Securing AI Systems, launched 15 October 2024, recommend a voluntary lifecycle approach to AI security covering development, deployment and operations.

    Confidence: Probable · As of 15 Oct 2024 · Evidenced
    Evidence and sources
    • Personal Data Protection Commission / IMDA T1 · Surfaced · retrieved 4 Nov 2026The Model Framework provides detailed and readily-implementable guidance to private sector organisations to address key ethical and governance issues when deploying AI solutions.

  • On 17 June 2026, CSA released for public consultation an addendum extending its 2024 AI security guidelines to agentic AI systems.

    Confidence: Confirmed · As of 17 Jun 2026 · Evidenced
    Evidence and sources
    • Cyber Security Agency of Singapore T1 · Retrieved · retrieved 4 Nov 2026CSA releases an addendum to support system owners in securing agentic AI systems for public consultation.

  • On 17 June 2026, the Cyber Security Agency of Singapore released, for public consultation, an addendum extending its AI security guidelines to agentic AI systems.

    Confidence: Confirmed · As of 17 Jun 2026 · Evidenced
    Evidence and sources
    • Cyber Security Agency of Singapore T1 · Retrieved · retrieved 4 Nov 2026CSA releases an addendum to support system owners in securing agentic AI systems for public consultation.

#

AI Verify is a voluntary, standardised technical testing framework and open-source toolkit, launched May 2022, letting organisations demonstrate responsible AI governance.

0 instruments (0 current) · 0 obligations · 3 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 79 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

AI Verify is a voluntary, standardised technical testing framework and open-source toolkit, launched in May 2022, that lets organisations demonstrate responsible AI governance. In October 2026 the AI Verify Foundation launched an AI Tester Accreditation Programme, accrediting a small number of third-party AI testing firms for a limited period on a voluntary basis; this is a private accreditation scheme and not a government-mandated conformity-assessment regime. The precise scope and duration of this programme remain uncertain in the available record.

Cited evidence in this module rests solely on tier 3/4 commentary despite primary official sources existing for this jurisdiction.

Claims and evidence (3)

  • AI Verify is a voluntary, standardised technical testing framework and open-source toolkit, launched May 2022, letting organisations demonstrate responsible AI governance.

    Confidence: Probable (capped: best source is T3) · Evidenced
    Evidence and sources
    • Future of Privacy Forum T3 · Retrieved · retrieved 4 Nov 2026AI Verify, an AI governance testing framework and toolkit, was launched in May 2022.

  • In October 2026 the AI Verify Foundation launched an AI Tester Accreditation Programme accrediting a small number of third-party testing firms for a limited period; this is a private accreditation and not a government-mandated conformity-assessment scheme.

    Confidence: Probable (capped: best source is T3) · As of 19 Oct 2026 · Evidenced
    Evidence and sources

  • The AI Verify Foundation launched an AI Tester Accreditation Programme in October 2026, accrediting a small number of third-party AI testing firms on a voluntary, non-statutory basis.

    Confidence: Uncertain (capped: best source is T3) · As of 19 Oct 2026 · Evidenced
    Evidence and sources

#

Elections (Integrity of Online Advertising) (Amendment) Act 2024 (Valid from 2025-01-22).

3 instruments (3 current) · 0 obligations · 7 claims · 0 milestones

Standing sub-brief · as of 12 Nov 2026 185 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Singapore has no general, tiered AI-law penalty regime; penalties that touch AI arise only from specific sectoral or criminal-law instruments. Under the Elections (Integrity of Online Advertising) (Amendment) Act 2024, breach of the election-deepfake prohibition carries individual criminal penalties, and a social media service that, without reasonable excuse, fails to comply with a corrective direction is separately liable on conviction to a fine of up to S$1,000,000. The Online Safety (Relief and Accountability) Act 2025 established the Online Safety Commission and is reported to have brought the first phase of statutory online-harm torts, including inauthentic material abuse covering AI-generated content, into force on 29 June 2026. The Criminal Law (Miscellaneous Amendments) Act 2025 criminalises producing or possessing non-consensual intimate images, including wholly AI-generated ones, with provisions commencing on 30 December 2025 and 17 August 2026. A large reported AI-chip diversion case in Singapore in 2026 — reportedly valued at approximately S$390 million and involving Nvidia-related hardware — is being prosecuted as fraud and money laundering rather than under the Strategic Goods (Control) Act 2002; this characterisation of the case is disputed in the available record.

Instruments (3)
  • Elections (Integrity of Online Advertising) (Amendment) Act 2024 Valid from 2025-01-22
    Binding
  • Online Safety (Relief and Accountability) Act 2025 Valid from 2026-06-29
    Binding
  • Strategic Goods (Control) Act 2002 In force since 2003-01-01
    Binding · Official text
Claims and evidence (7)

  • A large AI-chip diversion case (reported value approximately S$390 million) involving Nvidia-related hardware was reportedly being tried in Singapore as fraud and money laundering rather than as an export-control violation under the Strategic Goods (Control) Act 2002.

    Confidence: Disputed (capped: best source is T3) · Evidenced
    Evidence and sources
    • Model Diplomat T3 · Retrieved · retrieved 4 Nov 2026Singapore is trying its biggest AI-chip smuggling case not as an export-control violation, but as fraud and money laundering

  • Derived: Probable A large reported AI-chip diversion case in Singapore in 2026 is being prosecuted as fraud and money laundering rather than under the Strategic Goods (Control) Act 2002.

    Confidence: Probable (capped: best source is T3) · As of 1 Jul 2026 · Evidenced
    Evidence and sources
    • Model Diplomat T3 · Retrieved · retrieved 4 Nov 2026Singapore is trying its biggest AI-chip smuggling case not as an export-control violation, but as fraud and money laundering

  • Derived: Probable A social media service that, without reasonable excuse, fails to comply with a corrective direction in respect of an election-deepfake offence is liable on conviction to a fine of up to S$1,000,000.

    Confidence: Probable · As of 9 Sep 2024 · Evidenced
    Evidence and sources
    • Parliament of Singapore / Attorney-General's Chambers T1 · Surfaced · retrieved 4 Nov 2026Any person who is guilty of an offence under subsection (1) shall be liable on conviction by a District Court to a fine not exceeding $1,000 or to imprisonment for a term not exceeding 12 months or to both.

  • Singapore has no general tiered AI-law penalty regime; penalties that touch AI arise only from specific sectoral or criminal-law instruments.

    Confidence: Probable (capped: best source is T3) · As of 19 Oct 2026 · Evidenced
    Evidence and sources

  • Breach of the election deepfake prohibition carries individual criminal penalties and separate platform fines of up to S$1 million for non-compliance with takedown directions.

    Confidence: Probable (capped: best source is T3) · As of 22 Jan 2025 · Evidenced
    Evidence and sources
    • Resemble AI T3 · Retrieved · retrieved 10 Nov 2026Up to S$1,000 fine and 12 months’ imprisonment for individuals

  • OSRAA established the Online Safety Commission and brought the first phase of statutory online-harm torts, including inauthentic material abuse covering AI-generated content, into force on 29 June 2026.

    Confidence: Probable · As of 29 Jun 2026 · Evidenced
    Evidence and sources

  • The Criminal Law (Miscellaneous Amendments) Act 2025 criminalises producing or possessing non-consensual intimate images, including wholly AI-generated ones, with provisions commencing 30 December 2025 and 17 August 2026.

    Confidence: Probable · As of 17 Aug 2026 · Evidenced
    Evidence and sources
    • Ministry of Home Affairs (MHA), Singapore T1 · Retrieved · retrieved 10 Nov 2026The Criminal Law (Miscellaneous Amendments) Act 2025 ("the Act") was passed by Parliament on 4 November 2025.
    • AsiaOne T2 · Retrieved · retrieved 10 Nov 2026the definition of intimate images will now include materials that are generated to depict a person without altering an original image or recording.
    • Attorney-General's Chambers, Singapore (Singapore Statutes Online) T1 · Surfaced · retrieved 10 Nov 2026Sections 15, 16, 17, 24 to 28, 40 to 62, 65 to 71, 75 and 76 of the Criminal Law (Miscellaneous Amendments) Act 2025 come into operation on 17 August 2026.
    • The Sun (Malaysia) T2 · Retrieved · retrieved 10 Nov 2026the offence is punishable with mandatory imprisonment of up to two years, and the offender shall also be liable to a fine or caning

#

IMDA runs a Generative AI Evaluation Sandbox under its National AI Strategy 2.0 remit for organisations to test generative AI systems.

0 instruments (0 current) · 0 obligations · 1 claim · 0 milestones

Standing sub-brief · as of 12 Nov 2026 53 words · last cycle 12 Nov 2026 · AI research, not expert reviewed

Available reporting, assessed as uncertain, indicates that IMDA runs a Generative AI Evaluation Sandbox under its National AI Strategy 2.0 remit, allowing organisations to test generative AI systems. This is the only sandbox-related instrument on record for Singapore, and its precise scope and operational status are not clearly established in the available record.

Cited evidence in this module rests solely on tier 3/4 commentary despite primary official sources existing for this jurisdiction.

Claims and evidence (1)

  • IMDA runs a Generative AI Evaluation Sandbox under its National AI Strategy 2.0 remit for organisations to test generative AI systems.

    Confidence: Uncertain (capped: best source is T3) · As of 1 May 2026 · Evidenced
    Evidence and sources
    • AI Law Guide T3 · Retrieved · retrieved 4 Nov 2026It co-issues the Model AI Governance Frameworks (with PDPC and AI Verify Foundation), runs the Generative AI Evaluation Sandbox, drives the National Multimodal LLM Programme

#

Nothing recorded for this module in the research of 2026-11-12.

0 instruments · 0 obligations · 0 claims

No modules match.

Filters combine as OR inside a group and AND across groups.

Authorities

No authorities recorded in this research.

Enforcement actions

No enforcement actions recorded in this research.

Cases

No cases recorded in this research.

Consultations

ConsultationStatusOpened
Securing Agentic AI - Addendum to CSA's Guidelines and Companion Guide on Securing AI SystemsOpen
Securing Agentic AI - Addendum to the Guidelines and Companion Guide on Securing AI SystemsOpen

What this record does not yet establish is set out on the transparency register; its dated events are on the regulatory calendar.

Think something on this page is wrong? Report an error.

Technical provenance (record schema, run ID & build metadata)
Schema version
artificial-intelligence-record-v2.12
Research run ID
aic-union-SG-2026-11-04
Page model
module-card-page-v1
Renderer
render_jid_aic_v13
Build date