Artificial intelligence (AI) is already making decisions that impact Filipinos—in hiring, credit scoring, content moderation, and public services. Across Southeast Asia, governments are racing to put binding rules around AI. While the Philippines is actively moving in the same direction, its efforts remain scattered across agencies with overlapping mandates.
ASEAN guidelines exist, but adoption and enforcement depend on the discretion and capacities of member states
ASEAN’s regional response to AI includes two governance guidelines—the ASEAN Guide on AI Governance and Ethics in 2024 and its expanded guidance for Generative AI in 2025— developed under a broader digital cooperation agenda that also encompasses the Hanoi Digital Declaration adopted in January 2026 and the ASEAN Digital Masterplan 2030.
However, these documents remain non-binding, leaving implementation largely to the discretion of member states. This soft-law approach gives countries flexibility, given uneven priorities, capacities, and AI adoption. But it also creates the risk of people, firms, and public institutions facing different levels of protection and accountability depending on how each country governs AI.
As member states develop their own frameworks, national strategies are shaped by domestic priorities and institutional capacity, and also by the European Union (EU) as a regulatory superpower and the United States (US) and China as dominant AI innovators.
Global superpowers offer the foundational regulatory models that Southeast Asian nations look to as reference points
This external influence is clearest in how ASEAN states adopt AI systems, platforms, and standards developed elsewhere, making governance a task of translating EU, US, and Chinese models into local institutions, capacity, and priorities.
The EU, US, and China each approach governance differently.
The EU, for instance, has positioned itself as the leading model for binding AI regulation through the EU AI Act, widely regarded as the world’s first comprehensive and binding AI law. This Act uses a risk-based framework that prohibits certain uses and imposes strict obligations on high-risk applications, anchoring innovation in public trust, safety, and human rights.
While the EU emphasizes binding regional regulation, the US shapes AI governance through a more decentralized and innovation-oriented approach. Home to frontier AI firms like OpenAI, the US gives tech firms room to innovate while shaping global technology standards through federal agencies such as the NIST, sector regulators, and executive guidance.
China, meanwhile, pursues state-centered, application-specific control through targeted rules and standards rather than comprehensive legislation. For ASEAN, China matters as both a technology source and a model of direct state control over data, platforms, content, and digital infrastructure.
But differences in domestic capacity prevent the region from adapting to frontier, external models in a uniform way
Against these external models, ASEAN countries share risk-based language on lifecycle governance, ethics, accountability, and responsible deployment, but diverge in whether they translate these principles into model frameworks, circulars, presidential regulations, draft bills, or comprehensive statutes.
Singapore remains the regional outlier because its high state capacity and mature digital economy allow it to govern through the National AI Strategy 2019, the National AI Strategy 2.0 in 2023, the 2026 NAIS update and Model AI Governance Frameworks for Generative AI and Agentic AI, rather than a single prescriptive statute.
Vietnam has moved the fastest on comprehensive statutory law after passing its Law on Artificial Intelligence, effective March 2026. The law uses a tiered, risk-based structure similar to the EU model, classifying AI systems by the severity of their potential impact and imposing corresponding obligations on developers and deployers.
Malaysia and Thailand are likewise moving toward comprehensive legislation. Malaysia established a National AI Office in 2024 for cross-sector AI policy alignment and is drafting an AI Governance Bill, while Thailand continues to refine its draft of Principles of the AI Law. Both draft frameworks are heavily modeled after the EU AI Act.
Indonesia currently sits closer to Singapore in structure, using its National AI Strategy 2020-2045 and 2023 AI ethics circular to manage AI through sectoral governance. However, it is also drafting legally binding presidential regulations for AI, with the goal of positioning itself as a “digitally sovereign AI enabler”.
That variation also leaves the rest of ASEAN in a thinner governance position. Brunei, Cambodia, Laos, Myanmar, and Timor-Leste have less developed public AI governance frameworks and weaker institutional capacity to turn regional principles into enforceable rules.
The real divide is between states able to experiment with flexible governance and states that may need clearer statutory anchors. In practice, ASEAN AI governance is still stronger at defining guardrails such as ethics, safety, accountability, and risk assessment than at building the public-sector capacity, talent, infrastructure, and implementation systems needed to enforce those rules and support responsible adoption.
Like its neighbors, the Philippines is working toward a comprehensive legal framework
The Philippines, alongside Malaysia, Thailand, and Indonesia, is undergoing the same transition seen in Vietnam, moving from voluntary strategies toward stronger, more comprehensive legal frameworks.
Given these regional developments, the Philippines fits ASEAN’s pattern of active but transitional AI governance. While no executive order or enabling statute has assigns a lead institution for AI development and regulation, at least on the level of national policy the DOST leads the National AI Strategy and has pushed the National AI Center for Research and Innovation. Previously, the DTI developed the National AI Strategy Roadmap 2.0 and pioneered the original Center for AI Research (CAIR). DepEd issued the Foundational Guidelines on AI in Basic Education and now leads E-CAIR after its transfer from DTI. DEPDev is finalizing an AI Governance Framework, while DICT and the Civil Service Commission have worked on rules for government AI use.
These disparate efforts have shifted attention to consolidate action, to Congress, where lawmakers are considering AI bills for a legally binding, comprehensive AI governance framework. As of the current Congress, at least 50 AI-related bills and resolutions have been filed in both chambers, covering regulation, ethics, workforce impacts, education, and government use.
But the Philippines needs studies, capacity, and coordination to make any framework work
A single, comprehensive AI statute is not necessarily the only path forward. What matters more is whether existing and emerging frameworks are backed by risk-based studies and by institutional capacity, technical workforce, and implementation infrastructure needed to make them enforceable in practice. In middle powers like Japan, recent national laws on AI take on a light touch approach, focusing more on promotion of AI and establishing basic principles and frameworks, rather than imposing multi-tiered binding law such as the EU AI Act or Vietnam.
In the meantime, the more pressing need is coordination, which may require the designation of an existing agency to lead AI governance or establishing an inter-agency body to align overlapping mandates, reduce duplication, and ensure that national AI governance is coherent. So far, DOST has been taken to the task. But the Philippines will do better to consider policy options other than kitchen sinking or copying-pasting policies and controls into a comprehensive monolithic law. For the Philippines, the goal should not be creating the perfect framework but to ensure that any governance framework is wisely built and that the government is equipped to implement the plans it has committed itself to.
Regina Carla Bayongan is a Fellow under the first cohort of the AI Governance Accelerator Program of AI Safety Diliman, an organization focused on mitigating catastrophic risks associated with AI. A Political Science graduate of the University of the Philippines Diliman and a legislative researcher at the House of Representatives, she applies her background in public policy and international relations to the study of AI governance in the Philippines and Southeast Asia.
Josh Huesca is an AI Governance Scholar at AI Safety Diliman and a senior pursuing a BS in Health Sciences with a Minor in Management at the Ateneo de Manila University. His interests sit at the intersection of strategy, technology, and finance, with a focus on strengthening institutional responses that shape health, environmental outcomes, and public life.
Atty. Edsel Tupaz is Lead Researcher and Fellow at the AI Governance Accelerator Program of AI Safety Diliman. He is Senior Partner at Gorriceta Africa Cauton & Saavedra and leads the Data Privacy, Cybersecurity, and & AI Initiatives practice. He is included in the Top 100 Lawyers in the Philippines for 2023, 2024, and 2025.
The authors acknowledge the supervision of Ze Shen Chin of AI Standards Lab and a Research Affiliate at the Oxford Martin School AI Governance Initiative, as well as Lenz Dagohoy and Lexley Villasis of AI Safety Diliman.
AI Safety Diliman is supported by Kairos, an AI safety field-building nonprofit focused on strengthening the global AI safety and policy talent pipeline, and is fiscally sponsored by the Berkeley Existential Risk Initiative, an entity that supports academic and field-building efforts to reduce catastrophic risks associated with advanced technologies.