Cambodia is finalising its first National AI Strategy to promote ethical and transparent use of artificial intelligence amid growing concerns over unregulated systems, limited data protection laws and a lack of clear legal frameworks to support responsible innovation
This AI-generated image provided by Adobe shows a hummingbird. Artificial intelligence tools that can conjure whimsical artwork or realistic-looking images from written commands started wowing the public in 2022. (Adobe via AP)
Cambodia is finalising its first National Artificial Intelligence Strategy, as pressure grows to regulate the fast-developing technology in a country with no standalone AI laws and limited data protection.
The Ministry of Post and Telecommunications (MPTC) said the draft strategy will set ethical standards, transparency rules and oversight mechanisms. It follows a technical workshop in April and inter-ministerial consultations in May.
A national consultation workshop is planned for early July, with public feedback received until June 20. “The draft strategy is undergoing extensive review with input from government institutions, the private sector, academia and civil society,” MPTC said in a statement.
While the government has signalled interest in digital transformation, currently AI governance remains unregulated in Cambodia. The long-awaited cybercrime bill remains on pause, and the draft personal data protection law is still under review.
As a result, AI systems in use today are largely unregulated. Companies and agencies are not legally required to disclose how their systems make decisions, what data they collect or how it is used.
Hin Sopheap, Executive Director of DG Academy, said the risks of misuse are high in such an environment. He said, “We openly declare that our chatbot is built on a closed-source model from OpenAI and trained with our own and public data. But not everyone does this.”
Sopheap added, “Bad in, bad out, simple as that. If you train with biased data, you’ll get biased outcomes.”
He also cautioned against overregulation. “Cambodia needs to protect users, but also support startups. Most of us can’t build models from scratch. We rely on tools from global providers, like AWS or Azure.”
Sok Vanseka, Managing Partner of Sethalay Law and Director at the Credit Bureau of Cambodia, said the absence of clear AI rules poses challenges, especially in finance.
“Banks often don’t share data. It’s all isolated,” she said. “A collective approach, maybe led by the credit bureau, could help create more reliable datasets.”
She added that new initiatives, such as Cambodia’s cross-border credit reporting partnership with South Korea, show how fast the data landscape is evolving. “We need proper legal frameworks before we fully tap into these tools.”
Cambodia is still in the early stages of AI development. Its Digital Economy and Society Policy Framework 2021–2035 aims to guide the shift towards a more digitally-driven economy, but does not provide sector-specific rules for AI.
Chanty Pisal, Director of Digital Policy Research at the Cambodia Academy of Digital Technology (CADT), said, “Cambodia is in the early stages, but the foundation is promising.”
He pointed to UNESCO’s Ethics of AI Readiness Assessment, which conducted research between November 2024 and March 2025, surveying 26 ministries and institutions and gathering more than 120 stakeholder inputs.
Pisal said Cambodia still lacks quality data, reliable infrastructure and skilled talent. Adoption remains low, but efforts are underway, such as draft laws on data protection and new AI training programmes.
The MPTC is also contributing to ASEAN-level initiatives, including the ASEAN Guide on AI Governance and Ethics.
“There are lessons we can adapt,” Pisal said. “Sandbox approaches from Singapore and Malaysia show how flexible regulation can support innovation without compromising safety.”
For now, Cambodia is taking a voluntary, principle-based approach to AI governance. Officials describe it as “light-touch”, designed to foster innovation while promoting ethical use. But key gaps remain.
“We need clear direction to guide investment, address shortages and align development with national needs,” Pisal said. He added that the government is also working to finalise laws on personal data, cybersecurity and open data seen as critical for a functional AI ecosystem.