MDDI's Response to PQ on Reporting of Serious and Near-miss Incidents Involving Autonomous AI Systems

Parliament Sitting on 4 November 2026

Question for Oral Answer

*39. Mr Melvin Yong Yik Chye asked the Minister for Digital Development and Information whether the Ministry will consider introducing mandatory reporting requirements for serious incidents and near-misses involving autonomous AI systems, including unintended actions that could compromise sensitive data, disrupt essential services or cause significant public harm.

*40. Mr Sharael Taha asked the Minister for Digital Development and Information (a) whether existing cybersecurity reporting frameworks adequately capture near misses involving autonomous AI agents, including unintended actions and attempted unauthorised access; (b) whether a standardised near-miss reporting framework is being considered; and (c) how intelligence from such incidents can be shared across agencies and critical infrastructure operators to strengthen preventive safeguards.

Answer

Singapore has established reporting requirements for serious incidents involving personal data, cybersecurity and disruptions to critical or regulated services. These requirements apply regardless of whether AI contributed to the incident.

However, increasingly autonomous AI systems may take unintended actions that pose risks to safety and security. Such incidents may fall outside existing reporting thresholds, particularly when they do not result in reportable harm. Information on these incidents and near-misses could nevertheless provide valuable insights on emerging risks and help strengthen preventive safeguards.

IMDA and CSA are therefore studying how to strengthen the reporting of serious AI-related incidents, including whether additional reporting or information-sharing arrangements are needed. We will draw on international approaches and build on existing mechanisms, such as CSA’s exchange of cybersecurity intelligence with major technology providers and critical information infrastructure operators where feasible. Any additional arrangements should be targeted and proportionate, while taking into account legitimate security and commercial sensitivities.

International cooperation is also important given the cross-border nature of AI risks. Singapore will work with interested partners to develop a common understanding of AI incidents and near-misses, and explore practical arrangements for sharing relevant information securely.

*Converted to written answer