Opinion: Zac Cheah, co-founder of Pundi AI
The West regulates itself irrelevantly. As Europe and the US are stuck with committee meetings and draft laws, Southeast Asia, particularly Singapore, run live AI pilots in hospitals, refine crypto licensing through targeted enforcement, attracting top global talent with a working governance model.
Singapore’s secret? A sandbox-first approach that treats innovation not as a threat, but as an opportunity to carefully test it.
Architecture of Failure
The EU’s artificial intelligence law is an obvious case study. After years of discussion, we have created comprehensive regulations that companies face important compliance hurdles in their implementation, particularly in the gradual rollout timeline of the law. This has led to delays in adoption, particularly in healthcare and finance, and clarity is mission-critical.
The US isn’t even better. In 2024, more than 40 states introduced AI bills, but there was no federal framework to coordinate conflicting requirements. result? chaos. What is permitted in California may be prohibited in Texas. The fundamental problem is systematic. European and American regulators share a fundamental miscalculation that all theoretical risks must be eliminated before real-world innovation can be permitted.
The edge cases spent on monthly discussions are another month, where Singapore spends deploying AI systems, attracting talent and building irreversible strategic advantages.
Singapore’s sandbox revolution
Singapore has abandoned its first model of regulation in favour of real-world deployment under strict regulatory containment. The sandbox enables controlled, real-world testing with forced emergency shutdown protocols, layered fail-safes, and continuous compliance monitoring.
When Singapore’s Monetary Authority (MAS) observed crypto companies that fled Western uncertainty in 2024, it doubled the number of license approvals year-on-year. However, the recent evolution of Singapore’s regulations tells a more refined story. In June 2025, MAS placed the critical deadline required for locally established crypto companies that only serve overseas markets to obtain or suspend appropriate licenses. This is not a blanket crackdown, but rather a surgical enforcement targeting regulatory rulings.
Related: Bitstamp has granted a MAS license to operate in Singapore
The move is specifically addressed and utilised purely in Singapore, serving foreign customers without proper supervision. Companies faced a choice: commit to Singapore’s regulatory framework or exit.
Many chose to move rather than be exposed to proper surveillance. This exposed the number of things that use Singapore as a regulatory window dressing rather than a genuine operational base.
This enforcement measure indicates the maturity of the regulations in action. Singapore first built a legal infrastructure, endorsing 19 major cryptocurrency service providers, and then eliminated bad actors who exploited regulatory gaps. result? A high-quality crypto ecosystem with clear rules and serious players while competitors face ongoing regulatory disruptions.
Critics call this experiment, but Singapore has adopted a controlled approach. Each deployment caps user exposure, requires real-time data sharing, and includes an instant fallback system. This is not deregulation. Learning from reality rather than theory is agile and evidence-driven governance.
Payoff? This disciplined flexibility creates measurable returns. Singapore is currently the dominant AI hub in Southeast Asia, attracting global venture capital, world-class researchers and AI startups through its lucrative visa policies, strong research funding and strong industry partnerships. Its sandbox strategy is more than a regulatory experiment. This is a national advantage that transforms agility into a long-term competitiveness.
The fantasy of catch-up
Although Western perception is growing, implementation remains sluggish. As of mid-2025, the UK sandbox programme was in its early stages, with only a small cohort completed from the Financial Conduct Authority. In the United States, at the federal level, rulemaking often takes years from proposals to final rules, to final rules, including long public comments and interagency review phases. Meanwhile, state-level AI laws continue to multiply faster than cohesive federal approaches can manage.
This delay is not neutral. It’s economically destructive. By 2030, AI could contribute roughly $23 trillion to global GDP, but the lion’s share of its value will not be distributed evenly. Countries with an agile governance framework are systematically positioning themselves to grasp the majority of these benefits, with movers having far fewer economic opportunities.
The last tick of the watch
The message is clear. Singapore cleans homes, cracks down on regulatory rulings, supports a strong framework of serious operators, and deploys real-time AI across critical infrastructure. Code enforcement in June 2025 was not a retreat. The lack of sophistication implemented by Western regulators was ecosystem improvements. In this race, both speed and accuracy of regulation are forms of competitive advantage.
The Western economy has months rather than years to abandon its policy-paralytic approach and embrace evidence-based governance. Even at the grassroots level, the benefits of Singapore are exacerbated. The global AI race is accelerating, and like financial hubs, AI hubs are quickly emerging, focusing on policy, talent, access and competitive stakeholders.
Opinion: Zac Cheah, co-founder of Pundi AI.
This article is for general informational purposes and is not intended to be considered legal or investment advice, and should not be done. The views, thoughts and opinions expressed here are the authors alone and do not necessarily reflect or express Cointregraph’s views and opinions.
