Written Answer · 2026-09-08 · Parliament 15
Mandating Baseline Transparency Disclosures for Sensitive Applications of AI in Healthcare, Children and Personal Data
Ms Cassandra Lee asked the Minister for Digital Development and Information in a written question whether, given that IMDA's recently released Transparency Guidelines for Generative AI Chatbots are voluntary, the Government is considering mandating baseline transparency disclosures for sensitive AI applications such as healthcare, children or personal data, and what factors would inform that decision. Minister for Digital Development and Information Mrs Josephine Teo replied that the guidelines set a voluntary baseline for organisations to explain their generative AI chatbots' capabilities, limitations and safeguards. To address specific risks, the Government will consider factors such as how the technology is deployed, the nature of harms that may arise and the effectiveness of existing measures, as set out in the Ministry's 7 July 2026 reply to Mr Alex Yeo on regulating high-risk AI deployments. The reply did not say whether disclosures would be made mandatory.
Why it matters
IMDA's transparency guidelines for generative AI chatbots remain voluntary, and the Government did not commit to mandatory disclosure in sensitive areas such as healthcare and children.
Key Points
- • IMDA's Transparency Guidelines for Generative AI Chatbots set a voluntary baseline for organisations to explain chatbots' capabilities, limitations and safeguards.
- • To address specific risks, the Government will consider how the technology is deployed, the nature of harms that may arise and the effectiveness of existing measures.
- • These factors were set out in MDDI's 7 July 2026 written reply to Mr Alex Yeo on regulating high-risk AI deployments.
- • The reply made no commitment on mandating transparency disclosures for sensitive uses such as healthcare, children or personal data.
The Government keeps voluntary guidelines as the transparency baseline and will weigh further, risk-specific measures by deployment context, nature of harm and effectiveness of existing measures. It did not commit to mandatory transparency disclosures in healthcare, children or personal data.
Questioner Ms Cassandra Lee was concerned that voluntary guidelines may not be enough for sensitive AI uses, and asked the Government to consider mandatory baseline disclosures and explain what would drive that decision.
Singapore is sticking with a voluntary-first, risk-targeted approach to AI governance: mandatory rules will be considered only where specific, identifiable harms arise and existing measures fall short.
"In addition, to address specific risks, the Government will consider further factors, such as how the technology is deployed, the nature of harms that may arise, and the effectiveness of existing measures."
Participants (2)
Original Text (English)
SPRS Hansard · Fetched: 2026-09-26
39 Ms Cassandra Lee asked the Minister for Digital Development and Information in view of the fact that the Transparency Guidelines for Generative AI Chatbots recently released by the Infocomm Media Development Authority are voluntary (a) whether the Government is considering mandating baseline transparency disclosures for sensitive applications of AI, for example in healthcare, children or personal data; and (b) what factors will inform such a decision.
Mrs Josephine Teo : The Infocomm Media Development Authority's Transparency Guidelines for Generative AI Chatbots establish a voluntary baseline for organisations to explain the capabilities, limitations and safeguards of their generative AI chatbots.
In addition, to address specific risks, the Government will consider further factors, such as how the technology is deployed, the nature of harms that may arise, and the effectiveness of existing measures. These factors were mentioned in the Ministry of Digital Development and Information's reply to Mr Alex Yeo's Parliamentary question about the regulation of high-risk AI deployments on 7 July 2026. [ Please refer to "Key Risk Thresholds to Regulate High-risk AI Deployments and Implementing Mandatory Human Oversight for Fully Automated Decisions", Official Report, 7 July 2026, Vol 96, Issue 32, Written Answers to Questions section. ]