Submit Article
Legal Analysis. Regulatory Intelligence. Jurisprudence.
Singapore

Health Products (AI Standalone Mobile Application — Exemption) Order 2026

Overview of the Health Products (AI Standalone Mobile Application — Exemption) Order 2026, Singapore sl.

Statute Details

  • Title: Health Products (AI Standalone Mobile Application — Exemption) Order 2026
  • Act Code: HPA2007-S66-2026
  • Legislative Type: Subsidiary legislation (SL)
  • Authorising Act: Health Products Act 2007
  • Enacting Authority: Health Sciences Authority (HSA)
  • Enacting Power: Section 70 of the Health Products Act 2007
  • Citation: SL 66/2026
  • Commencement: 13 February 2026
  • Status: Current version as at 27 March 2026
  • Key Provisions:
    • Section 2: Definitions
    • Section 3: Exemption from manufacturer’s licence requirements (public healthcare entities)
    • Section 4: Exemption from registration requirements (public healthcare entities supplying AI standalone mobile applications)

What Is This Legislation About?

The Health Products (AI Standalone Mobile Application — Exemption) Order 2026 (“the Order”) is a targeted regulatory exemption under Singapore’s Health Products Act 2007. In essence, it allows certain public healthcare entities to develop, supply, and deploy AI standalone mobile applications that qualify as low to moderately low risk medical devices, without having to comply with two specific regulatory burdens: (i) the manufacturer’s licence requirement and (ii) the medical device registration requirement.

The Order recognises that public healthcare institutions and related entities may need to innovate quickly—particularly when AI tools are used to support diagnosis, treatment, or testing within healthcare services. Rather than requiring every internally developed AI app to go through the full licensing and registration pathway applicable to manufacturers generally, the Order creates a controlled “fast track” for qualifying AI apps, while still imposing safeguards such as clinician supervision, notification to the Authority, and internal endorsement by senior clinical governance leadership.

Practically, the Order is designed to support the deployment of AI-enabled mobile medical devices within the public healthcare ecosystem (including cluster-related entities and specified organisations), while maintaining regulatory oversight through risk classification, governance controls, and pre-deployment endorsement.

What Are the Key Provisions?

1. Definitions and the scope of “AI standalone mobile application”

Section 2 defines an “AI standalone mobile application” as a medical device that is an artificial intelligence software and a standalone mobile application as defined in the Health Products (Medical Devices) Regulations 2010. This is important because the exemption is not for ordinary software; it applies only where the AI mobile application is treated as a medical device under the Health Products Act 2007 framework.

The Order also defines “public healthcare entity” broadly. It includes: (a) a public healthcare institution, (b) a cluster HQ, (c) Synapxe Pte. Ltd., and (d) MOH Office for Healthcare Transformation Pte. Ltd. The definition of “public healthcare institution” is tied to entities that hold a licence under the Healthcare Services Act 2020 (HCSA), including subsidiaries of cluster HQs, businesses registered by cluster HQs, and entities managed and controlled by a cluster HQ.

2. Exemption from manufacturer’s licence requirements (Section 3)

Section 3 provides that Section 12(1) of the Health Products Act 2007 does not apply to a public healthcare entity X that manufactures an AI standalone mobile application, provided all conditions are met.

The conditions are cumulative:

  • Risk class: the AI standalone mobile application must fall under Class A (low risk) or Class B (moderately low risk) of the Third Schedule to the Health Products (Medical Devices) Regulations 2010.
  • Clinician supervision: manufacture must be under the supervision of a clinician employed in a public healthcare institution and holding the position of consultant or higher.
  • Intended use within licensed healthcare services: the app must be intended for use either:
    • by X (if X holds an HCSA licence) to provide a healthcare service that X is licensed to provide; or
    • by another public healthcare entity Y (or by HPB) to provide a healthcare service that Y is licensed to provide.
  • Notification to the Authority: X must notify the Authority of its manufacture of the AI standalone mobile application prior to or at any time during manufacture.
  • Joint manufacturing: if the AI app is jointly manufactured with another person, that other person must be a public healthcare entity.

Legal significance: Section 3 does not remove all regulatory duties; it specifically exempts qualifying public healthcare entities from the manufacturer’s licence requirement. However, the exemption is conditional on risk classification, governance, and notification. For practitioners, this means that any deviation—such as manufacturing outside the supervision requirement, using the app for an unlicensed service, or failing to notify—could jeopardise the exemption.

3. Exemption from registration requirements (Section 4)

Section 4 addresses a separate regulatory obligation: the registration requirement in Section 15(1) of the Act. It provides exemptions where a public healthcare entity X supplies an AI standalone mobile application, whether manufactured by X or by other public healthcare entities, again subject to strict conditions.

Section 4(1): If X supplies an AI standalone mobile application that X manufactures (alone or jointly), Section 15(1) does not apply if the “common conditions” in Section 4(3) are satisfied.

Section 4(2): If X supplies an AI standalone mobile application that X does not manufacture, Section 15(1) also does not apply, provided that the app is manufactured by one or more other public healthcare entities and the common conditions are satisfied.

Section 4(3): the common conditions include:

  • Risk class: again, Class A or Class B.
  • Clinician supervision: manufacture under supervision of a consultant-or-higher clinician employed in a public healthcare institution.
  • Permitted supply pathways: X may supply the app:
    • by way of administration to or application in any person in the course of diagnosis, treatment or test, if X holds an HCSA licence to provide such healthcare service; or
    • to another public healthcare entity Y for Y to supply to another licensed public healthcare entity or to HPB; or
    • directly to HPB for HPB to provide a licensed healthcare service.
  • Notification of deployment: X must notify the Authority of X’s or Z’s deployment (as applicable) prior to deployment.
  • Endorsement for use: prior to deployment, the AI app must be endorsed by senior leadership, with the specific endorser depending on whether the entity has a medical board and/or a chief executive officer, or if the entity is HPB (endorsed by the Chief Executive of HPB).

4. Endorsement mechanics (Section 4(3)(e))

The endorsement requirement is a governance safeguard. It is structured as a conditional matrix:

  • If the entity has both a medical board and a chief executive officer: endorsement by either the chairman of the medical board or the chief executive officer.
  • If it has a medical board but no chief executive officer: endorsement by the chairman of the medical board.
  • If it has a chief executive officer but no medical board: endorsement by the chief executive officer.
  • If the entity is HPB: endorsement by the Chief Executive of HPB.

For counsel, this is a practical compliance checkpoint: the exemption from registration is not available unless endorsement is obtained before deployment in accordance with the entity’s governance structure.

How Is This Legislation Structured?

The Order is short and structured around four provisions:

  • Section 1 (Citation and commencement): provides the name of the Order and its commencement date (13 February 2026).
  • Section 2 (Definitions): sets out key terms, including “AI standalone mobile application,” “public healthcare entity,” “cluster HQ,” and references to the relevant Regulations and Acts.
  • Section 3 (Exemption from manufacturer’s licence requirements): exempts qualifying public healthcare entities from the manufacturer’s licence requirement when manufacturing AI standalone mobile applications under specified conditions.
  • Section 4 (Exemption from registration requirements): exempts qualifying public healthcare entities from the device registration requirement when supplying AI standalone mobile applications, whether manufactured by themselves or other public healthcare entities, subject to common conditions including risk class, supervision, notification, and pre-deployment endorsement.

Who Does This Legislation Apply To?

The Order applies to public healthcare entities as defined in Section 2. This includes public healthcare institutions licensed under HCSA, cluster HQs, Synapxe Pte. Ltd., and MOH Office for Healthcare Transformation Pte. Ltd. In addition, the exemption is operationally linked to whether the entity holds an HCSA licence to provide the relevant healthcare service.

It also indirectly involves other parties only to the extent that joint manufacturing is permitted only with other public healthcare entities (Section 3(e)), and supply pathways are limited to other public healthcare entities and/or HPB (Section 4(3)(c)). Private manufacturers or non-public entities are therefore outside the exemption’s intended perimeter, unless they fall within the definition of “public healthcare entity.”

Why Is This Legislation Important?

This Order is significant because it creates a regulatory carve-out for a specific category of AI-enabled medical devices: AI standalone mobile applications that are low to moderately low risk and developed within the public healthcare governance framework. By exempting qualifying entities from manufacturer licensing and device registration, it reduces time and administrative burden—supporting faster translation of AI tools into clinical use.

At the same time, the Order is not a blanket exemption. It embeds safeguards that practitioners should treat as essential compliance elements: risk classification (Class A/B), clinician supervision by consultant-or-higher clinicians, notification to the Authority (both at manufacture and prior to deployment), and formal endorsement by senior clinical governance leadership before deployment.

From an enforcement and risk-management perspective, the conditional nature of the exemptions means that legal teams should implement internal processes to verify: (i) the device classification under the Third Schedule; (ii) the clinician supervision arrangements; (iii) the intended clinical use aligns with HCSA-licensed services; (iv) timely notifications are made; and (v) endorsement is obtained by the correct office-holder(s) depending on the entity’s governance structure.

  • Health Products Act 2007 (including Sections 12(1) and 15(1), and the exemption-making power in Section 70)
  • Health Products (Medical Devices) Regulations 2010 (including Regulation 26(5) and the Third Schedule risk classes)
  • Healthcare Services Act 2020 (definition of “healthcare service” and licensing framework)
  • Health Promotion Board Act 2001 (establishing HPB)
  • Business Names Registration Act 2014 (relevant to the definition of “public healthcare institution”)

Source Documents

This article provides an overview of the Health Products (AI Standalone Mobile Application — Exemption) Order 2026 for legal research and educational purposes. It does not constitute legal advice. Readers should consult the official text for authoritative provisions.

Written by Sushant Shukla

More in

Legal Wires

Legal Wires

Stay ahead of the legal curve. Get expert analysis and regulatory updates natively delivered to your inbox.

Success! Please check your inbox and click the link to confirm your subscription.