Singapore tightens AI-cybersecurity rules, mandating top Cyber Trust Mark by 2027 and building AI agents registry; co-governance emerges as new model for critical infrastructure

Singapore is enhancing its cybersecurity framework by mandating that Critical Information Infrastructure (CII) owners achieve the highest tier of the Cyber Trust Mark by 2027. This initiative aims to bolster cyber resilience and address supply chain vulnerabilities in an era where AI is increasingly integral to infrastructure.¹²

The Cyber Security Agency of Singapore's proactive measures, including the establishment of an AI agents registry, signify a shift towards co-governance as a model for managing systemic risks. This registry, developed through a regulatory sandbox, provides structured visibility into the ownership and function of AI agents across 150,000 public officers, fostering trust without compromising proprietary systems.³

As AI technologies evolve, the threat landscape has transformed, with adversaries exploiting vulnerabilities at machine speed. “Frontier AI has fundamentally changed the character of the threat environment,” highlighting the need for real-time cybersecurity measures. Traditional governance frameworks, designed for slower threats, are becoming obsolete.

To address these challenges, Singapore emphasizes the importance of early risk identification and incident simulations to prepare for potential AI-driven failures. “Co-governance allows for the real-time sharing of threat intelligence,” shifting public cybersecurity from a reactive to a proactive stance. Despite these advancements, Singapore acknowledges the need for ongoing collaboration to manage evolving AI supply chains and cross-border dependencies.

In summary, Singapore's approach reflects a comprehensive strategy to secure AI-enabled infrastructure, ensuring that cybersecurity measures are integrated before deployment rather than as a reaction to failures.