Steps taken to help organisations adopt AI safely and responsibly: Forum

We thank Calvin Zhang Song for his letter and constructive suggestions (Set clear approval rules for AI-generated tools, Oct 7).

The writer raises important points on how organisations can safely use generative AI (GenAI) tools. The recent incident in which an employee unintentionally exposed e-mail addresses with an AI-generated script is a timely reminder that organisations must put in place clear rules and safeguards to manage risks, especially when personal data is involved.

To guide organisations, the Personal Data Protection Commission has issued an advisory on data protection risks when using unauthorised AI tools for software development and advisory guidelines on the use of personal data in GenAI.

Organisations remain responsible for protecting personal data, regardless of whether AI is used. They should establish clear policies and approval processes for AI tools, assess data protection risks, and review and test AI-generated code before deployment.

Testing should use dummy data or accounts and verify what users actually see, rather than rely solely on system logs. Organisations should also implement appropriate technical safeguards and train employees to recognise risks and seek specialist review when needed.

The Infocomm Media Development Authority recognises that smaller businesses may need more support, and has published resources such as the GenAI Playbook for Enterprises, with governance and security guidance for deploying AI in an enterprise context, as well as the Starter Kit for Testing LLM-Based Applications to help organisations assess AI safety and reliability. Organisations can also tap the upcoming AI Verify Foundation’s AI Tester Accreditation Programme, which will maintain a list of accredited firms to assess AI systems.

As AI adoption grows, organisations must strengthen governance and data protection practices to innovate confidently while maintaining customer trust.

Jo-Anne Tan
Director, Communications and Marketing
Infocomm Media Development Authority 

Source: The Straits Times © SPH Media Limited. Permission required for reproduction.

ADV: Global Rule of Law Conference - Fortifying the Rule of Law Against Future Challenges Your own personal AI agent. What could possibly go wrong?: Opinion