The Infocomm Media Development Authority (IMDA) of Singapore today launched a series of initiatives to safeguard generative AI and cross‑border data use at the inaugural Singapore Data Festival, unveiling consent and transparency guidelines that define how organisations may train and deploy AI on customer data, alongside new Privacy Enhancing Technologies (PETs) guides and Singapore’s first Digital Twin for Enterprises Playbook.
The event also saw Singapore’s Personal Data Protection Commission (PDPC) sign a Memorandum of Cooperation with Japan’s Personal Information Protection Commission to align data protection across borders as CIOs balance Gen AI scaling with staying compliant and preserving trust.
The new Advisory Guidelines on the use of Personal Data in Gen AI define how data can be collected and used across the AI lifecycle, including handling sensitive new inputs such as end‑user prompts and responding to individuals’ requests about their personal data in AI systems.
Mr Derrick Goh, Group Chief Operating Officer, DBS, said “Since embarking on our AI journey in 2014, DBS has scaled our AI use across the bank with a focus on customer experience underpinned by the application of responsible AI practices. We are pleased to contribute to IMDA’s Guidelines based on insights from our Gen AI initiatives such as our customer-facing innovations, DBS Joy and digibot. We welcome IMDA’s pragmatic approach to balancing transparency with implementation flexibility that will help strengthen trust in AI across the ecosystem and advance Singapore’s ambition to be a trusted AI financial hub.”
Enterprises training models on personal data will be expected to obtain specific consent, turning PDPA obligations into operational requirements that can be built into data pipelines and MLOps processes in both domestic and cross‑border contexts, and sit alongside PDPC’s new Federated Learning and updated Synthetic Data Generation guides that help organisations collaborate on data and generate artificial datasets without exposing raw records.
As more enterprises launch AI customer agents, IMDA’s Transparency Guidelines for Gen AI Chatbots push organisations to treat AI assistants as regulated customer channels, backed by “Chatbot Info Cards” that disclose capabilities, limits, safety practices, data use and escalation options at a glance.
These measures, together with IMDA’s Digital Twin for Enterprises Playbook and PET Sandbox use cases such as logistics players using live operational twins to predict failures and optimise flows, and hospitals analysing medical images securely in the cloud, aim to ensure that enterprise AI in Singapore will be built not just on powerful models, but on explicit governance, interoperable privacy frameworks and customer‑centric data transparency.