A reflection on why good regulation does not try to predict the entire future — it creates a disciplined way to learn before uncertainty becomes irreversible harm.
Traditional regulation assumes that rules can be written after a technology becomes sufficiently understood. Artificial intelligence changes too quickly for that sequence to work comfortably. Rules written too early may block useful innovation. Rules written too late may allow harmful practices to become deeply embedded.
The answer is not to eliminate regulation. It is to make regulation more capable of learning.
The OECD describes regulatory sandboxes and experimental governance as ways for authorities and innovators to test new systems under supervision. Its recent work on anticipatory governance also emphasises flexible frameworks, horizon scanning and early identification of emerging risks.
Thailand should develop regulatory learning environments in areas such as health AI, financial services, autonomous mobility, energy systems, educational technology and digital identity.
But a sandbox should not become a place where companies receive permission to experiment while affected citizens remain outside the process. Regulators, technology providers, domain experts, consumer representatives and frontline users should examine the system together.
The purpose is not simply to determine whether the technology works. It is to understand who benefits, who carries the risk, what evidence is required and what should happen when the system produces harm.
My research recommended consultation and reducing fear of contradicting leadership inside technology-mediated communities. Regulatory innovation requires the same culture. Officials must be able to challenge political urgency. Technical experts must be able to explain uncertainty. Smaller organisations should be able to raise concerns alongside major corporations.
Adaptive regulation does not mean constantly changing rules without direction. It means maintaining clear principles while improving practical requirements as evidence develops.
Thailand should aim to become neither the country that blocks every innovation nor the country that accepts every experiment. It should become a country capable of learning responsibly before making new systems permanent.
Good regulation does not try to predict the entire future. It creates a disciplined way to learn before uncertainty becomes irreversible harm.