Singapore's Personal Data Protection Commission is proposing mandatory AI-specific notifications for organisations that use personal data to train generative AI models.
This move aims to increase transparency for individuals whose information is used to develop artificial intelligence. As generative AI becomes more integrated into business operations, the risk of data misuse increases, necessitating clearer boundaries between general data collection and AI training.
The PDPC released the proposed guidelines on June 2, 2026 [1]. Under the proposal, firms would be required to issue notifications that explicitly explain how personal data will be used to develop generative AI systems. This represents a shift from previous practices where some organisations relied on broad privacy notices to cover the use of data for AI training [2].
The commission intends for these notices to ensure that individuals are fully aware of the specific purposes for which their data is being processed. By requiring AI-specific disclosures, the regulator seeks to protect personal data, and give users more clarity on the lifecycle of their information within an AI model.
Currently, there is a discrepancy in how firms handle these notifications. Some organisations continue to use general privacy policies, while others have already begun issuing specific notices regarding AI training [2]. The proposed guidelines would standardise this requirement across the board in Singapore.
The PDPC has not yet finalised the guidelines, but the proposal marks a significant step toward tighter regulation of the AI industry. The focus remains on balancing technological innovation with the fundamental right to data privacy.
“Singapore's PDPC is proposing that firms must issue AI‑specific privacy notices when using personal data to train generative AI models.”
This proposal signals a move away from 'catch-all' privacy policies toward granular consent in the age of generative AI. By mandating specific notifications, Singapore is establishing a regulatory precedent that treats AI training as a distinct data processing activity rather than a routine business operation, potentially influencing how other global financial and tech hubs approach AI governance.
