Singapore has unveiled a pioneering Model AI Governance Framework for Generative AI, aiming to strike a balance between innovation and safety in the rapidly evolving field of artificial intelligence. Released on May 30, the framework seeks to address the unique risks and challenges posed by generative AI, while fostering a trusted ecosystem for its development and deployment.

Generative AI, known for its ability to create text, images, and other media types, has captured global attention with its transformative potential. However, it also brings significant risks, such as bias, misinformation, lack of explainability, and new issues like hallucination and copyright infringement. The framework builds on Singapore’s earlier AI governance efforts, first introduced in 2019 and updated in 2020, expanding its scope to specifically address the nuances of generative AI.

The framework is structured around nine key dimensions:

• Accountability: Ensuring accountability throughout the AI development chain is critical. The framework highlights the importance of clear allocation of responsibilities among model developers, application deployers, and cloud service providers. Drawing parallels with established cloud computing models, it advocates for shared responsibility to protect end-users and ensure overall system security.

• Data: Data quality and ethical use are fundamental to AI model development. The framework emphasizes using trusted data sources and addressing contentious issues like personal data and copyright material pragmatically. Policymakers are encouraged to provide clarity on legal applications to facilitate the responsible use of data.

• Trusted Development and Deployment: Transparency and adherence to best practices in AI model development and deployment are crucial. The framework calls for industry-wide adoption of safety measures and the disclosure of essential information, akin to “food labels,” to enable informed decisions by users.

• Incident Reporting: Recognizing that no AI system is foolproof, the framework advocates for robust incident reporting structures. This includes timely notification and remediation processes, learning from established practices in domains like telecommunications and finance.

• Testing and Assurance: Third-party testing and assurance play a vital role in building trust. The framework promotes the adoption of standardized testing methodologies and independent verification to ensure consistent and reliable AI performance.

• Security: Generative AI introduces new security risks, necessitating the adaptation of traditional information security frameworks. The framework recommends developing new security tools and practices tailored to the unique characteristics of AI.

• Content Provenance: To combat misinformation and other harms, the framework supports the implementation of technical solutions like digital watermarking and cryptographic provenance. These measures aim to enhance transparency about the origin of AI-generated content.

• Safety and Alignment Research & Development (R&D): Accelerated investment in R&D is essential to improve AI model alignment with human values and intentions. The framework encourages global cooperation among AI safety institutes to optimize resources and address both current and future risks effectively.

• AI for Public Good: Beyond risk mitigation, the framework envisions AI as a tool for public benefit. This includes democratizing access to AI technologies, enhancing public sector adoption, upskilling workers, and promoting sustainable AI development.

The release of this framework is a significant step in shaping the future of AI governance. It is designed to evolve alongside technological advancements and policy discussions, ensuring that AI can be harnessed safely and effectively for the public good. By engaging key stakeholders, including policymakers, industry leaders, and the research community, Singapore aims to create a comprehensive and adaptable governance structure that addresses the multifaceted challenges of generative AI.

Need Help? 

If you’re wondering how Singapore’s AI framework, or any other framework, regulation or bill around the world, could impact you, don’t hesitate to reach out to BABL AI. Their Audit Experts are ready to provide valuable assistance while answering your questions and concerns.