The Deputy Prime Minister of Singapore, Heng Swee Keat, has launched the Model Governance Framework for GenAI at the Asia Tech x Singapore event. Developed by the AI Verify Foundation and IMDA, the framework addresses concerns related to GenAI, emphasizing transparency and ethical governance to combat misinformation.

It identifies nine key areas for strengthening governance, including accountability, trusted data for AI training, and content authenticity. The framework calls for clearer personal data laws and collaboration to curate relevant training data sets, especially for low-resource languages. It also suggests technical solutions like digital watermarking to verify AI-generated content.

The guidelines were established after consultation with over 70 tech-specialized organizations, including Microsoft, Google, and the US Department of Commerce. It also aligns with international AI principles and initiatives, such as the Hiroshima AI Process and the ASEAN Guide on AI Governance.

Analyst QuickTake: Singapore's new Governance Framework implies that firms must invest in enhancing their governance frameworks, processes, and tools to align with stated principles. Similar developments in other regions include EU approval of the AI Act in May 2024 , which regulates AI applications on a risk scale and prohibits unacceptable AI practices. In February 2024 , the Biden-Harris administration announced the first-ever consortium related to AI Safety in the US, which includes guidelines for red teaming, risk management, safety, and security.