ElendiLabs Logo
Back to Articles

Need Regulatory Help? Try Our Platform

Post your regulatory questions or request quotations from verified pharmaceutical consultants worldwide. Get matched with experts who specialize in your market.

Registration

December 16, 2025

Approximately 5 minutes

Singapore's SaMD Regulation: Cybersecurity, AI, and Compliance

Singapore SaMD Regulation overview showing definitions of Software as a Medical Device, IMDRF risk classifications Class A–D, and key regulatory focus areas including cybersecurity requirements, versioning & traceability, change management, and AI-powered device compliance.

Singapore's SaMD Regulation: Cybersecurity, AI, and Lifecycle Compliance

Software as a Medical Device (SaMD), which includes standalone software, web-based applications, mobile apps, and Artificial Intelligence (AI) solutions, is comprehensively regulated by the Health Sciences Authority (HSA) in Singapore. HSA employs a lifecycle approach, outlined in its Regulatory Guidelines for Software Medical Devices, covering development, registration, and post-market obligations.


Classification and Registration

The HSA follows the International Medical Device Regulators Forum (IMDRF) framework for classifying SaMD, which is based on the significance of the information provided by the software and the state of the healthcare situation or condition. SaMD is categorized into four risk classes (A, B, C, D).

SaMD Classification (IMDRF)Risk LevelExample SaMD Functions
Class ALow RiskProvide information that drives clinical management, but non-serious condition (e.g., patient education app).
Class BLow to Medium RiskProcess, analyze, or create information for clinical management in non-serious conditions (e.g., basic diagnostic image viewing).
Class CMedium to High RiskProcess or analyze information for clinical management in serious conditions (e.g., software suggesting treatment options for cancer).
Class DHigh RiskProvide critical information for clinical management in critical conditions or states (e.g., software that monitors patient data to recommend immediate life-saving intervention).

All registrable SaMD must undergo the standard product registration process, adhering to the requirements of the determined risk class.


Key Regulatory Focus Areas for SaMD

1. Cybersecurity Requirements

The HSA places a high emphasis on cybersecurity risk management throughout the SaMD lifecycle. Registration dossiers must be supported by a documented cybersecurity strategy, including:

  • Secure-by-Design Architecture: Integrating security measures from the initial development phase.
  • Threat Modeling: Systematic identification and assessment of potential vulnerabilities.
  • Vulnerability Assessments: Ongoing testing to detect and mitigate risks.
  • Incident Response Plans: Detailed plans for real-time threat detection and response in the post-market phase.

2. Versioning and Traceability

Clear and consistent software versioning is mandatory for proper identification and post-market traceability. Labeling requirements for SaMD (GN-23) specify that:

  • The software version number must be clearly displayed (e.g., on the splash screen or user interface for downloaded or web-based apps).
  • The versioning data must be submitted as part of the registration dossier and must reflect changes in functionality, user interface, or bug fixes.

3. Managing Changes (Change Notifications)

Any change to a registered SaMD requires a Change Notification to the HSA. Changes are classified based on their impact:

  • Significant Changes: Require a more rigorous technical review and include major algorithm modifications, introduction of new AI features, or interface redesigns that impact usability or safety.
  • Non-Significant Changes: Typically administrative or minor bug fixes that do not affect the intended use or risk profile.

4. AI-Powered Medical Devices (AI-MD)

AI-based SaMDs must comply with all medical device regulations and specific data privacy laws in Singapore, such as the Personal Data Protection Act (PDPA). The HSA's guidance (GL7) outlines principles for manufacturers implementing adaptive or continuously learning algorithms, emphasizing:

  • Addressing the regulatory implications of continuous learning models and model retraining.
  • Ensuring ongoing performance monitoring and collecting real-world evidence.
  • Periodic reporting to the HSA on AI model performance.

Ask Anything

We'll follow up with you personally.

100% response rate • Reply within 7 business days

Your email will not be published. We'll only use it to notify you when we respond.

Need Expert Guidance?

Contact us at contact@elendilabs.com / +852 4416 5550