Strengthen Rules Governing Use of Classified and Unclassified Government Data Given Rapid Development of AI Technologies

56 Mr Fadli Fawzi asked the Minister for Digital Development and Information given the rapid development of artificial intelligence (AI) technologies, what updates, if any, have there been to strengthen the rules governing the use of (i) classified and (ii) unclassified government data, respectively, in AI chatbots and platforms in the Public Service.

Mrs Josephine Teo: The Government's risk-calibrated approach to data security in artificial intelligence (AI) systems was explained in a written reply, given on 9 January 2024, to related questions asked by Dr Tan Wu Meng and Mr Gerald Giam. [Please refer to ​"Regulations on Input Prompts for Large Language Models to Prevent Disclosure of Confidential Data", Official Report, 9 January 2024, Vol 95, Issue 118, Written Answers to Questions for Oral Answer not Answered by End of Question Time section.]

That overarching approach continues to apply today. We mandate technical and contractual safeguards commensurate with the classification and sensitivity of data involved; these are complemented by responsible use policies and governance processes practiced by officers.

As AI technologies evolve, we continue to apply these same principles to new use cases, tools and capabilities. We regularly review our rules and safeguards to ensure agencies can harness AI to transform public service delivery while keeping Government data and systems secure.