Thailand beats global AI average but governance gaps remain

Thailand has ranked 60th among 135 countries and jurisdictions in the 2026 Global Index on Responsible AI, reflecting solid foundations in workforce readiness and infrastructure but persistent weaknesses in policy implementation, inclusion and public participation.

The country received an overall score of 36.99 out of 100, placing it in the middle of the global table. The score was slightly above the worldwide average of about 35 and comfortably higher than the Global South average of approximately 27, but remained well below the Global North average of around 55.

The Global Index on Responsible AI, or GIRAI, evaluates how effectively countries govern artificial intelligence and translate responsible-AI principles into practical policies, institutions and protections.

The 2026 edition covers evidence collected between November 1, 2023 and September 30, 2025, meaning more recent Thai policy developments are not reflected in the ranking.

Workforce and infrastructure provide a solid base

Thailand’s comparatively stronger results came from workforce readiness, institutional capacity and supporting infrastructure.

These areas indicate that the country has a reasonable foundation for adopting AI, developing related skills and supporting innovation.

The findings suggest that Thailand is not among the countries least prepared for responsible AI. Its existing policy vision, institutional structures and innovation potential place it ahead of several regional peers.

However, the report also points to a widening gap between Thailand’s ability to introduce AI and its capacity to ensure that those systems are transparent, accountable and safe for the public.

Civil-society participation remains a major weakness

Thailand’s lowest result was for civil-society organisation engagement, which received only 19.55 points.

The score reflects limited participation by civic groups, workers’ organisations, rights advocates, independent researchers and affected communities in the development and oversight of AI policy.

Thailand also scored only 29.57 points for AI policy, indicating that existing frameworks, enforcement mechanisms and implementation arrangements remain insufficiently developed.

The results highlight a substantial divide between having strategies and institutions on paper and ensuring that those structures deliver enforceable protection in practice.

The country’s responsible-AI framework remains weakened by uncertainty over enforcement, limited independent scrutiny and inadequate remedies for people affected by harmful or discriminatory AI systems.

Thailand ranks fifth among ASEAN countries assessed

Thailand placed fifth among the ASEAN countries covered by the index.

Singapore led the region with 45.21 points, followed by Vietnam with 41.56, Indonesia with 38.83 and Malaysia with 38.54.

Thailand’s score of 36.99 placed it ahead of the Philippines at 32.98, Cambodia at 24.09, Laos at 14.45 and Myanmar at 11.71.

The regional comparison places Thailand in ASEAN’s middle tier. The country has stronger foundations than several neighbours but has yet to reach the group of regional leaders.

Its position reflects the same pattern seen in the global results: relatively good readiness and infrastructure, but weaker safeguards, enforcement and public participation.

Six priorities for stronger AI governance

The report identifies six policy areas Thailand should advance to move from broad principles towards mechanisms that work in practice.

1. Register government AI systems

Thailand should establish a public register of AI systems used by state agencies.

Such a register would allow the public to see where AI is being deployed, what decisions it supports and which authorities are responsible for its operation.

2. Strengthen AI procurement standards

Government agencies should adopt clearer standards when purchasing or commissioning AI systems.

Procurement rules should require transparency, safety checks, accountability and evidence that the technology serves the public interest.

3. Expand civil-society participation

The authorities should provide formal channels for civic groups, affected communities and independent organisations to take part in AI policymaking and oversight.

The country’s low civil-society score makes this one of the most urgent areas for improvement.

4. Protect children, equality and diversity

Stronger safeguards are needed to protect children’s rights, promote equality and recognise Thailand’s linguistic and cultural diversity.

AI systems trained on incomplete or biased data can produce discriminatory results, particularly for groups already facing exclusion.

5. Protect workers while accelerating skills development

Thailand should combine reskilling and upskilling programmes with stronger protections for workers affected by automation.

Workforce development should not focus only on preparing people for new technology, but also on protecting employment rights and working conditions.

6. Include environmental impacts in AI governance

Thailand should incorporate environmental concerns into its AI policies.

These include electricity consumption, data-centre demand and the broader environmental costs of developing and operating AI systems.

Challenge is turning policy into protection

The index concludes that Thailand has promising foundations in policy direction, institutional structure and innovation capacity.

These strengths have kept the country above the global average and ahead of several ASEAN neighbours.

However, Thailand has not yet entered the leading group because enforcement remains unclear, independent oversight is limited and mechanisms for compensating or assisting people harmed by AI are underdeveloped.

The country’s next step is therefore not simply to publish more strategies or principles.

Its main challenge is to ensure that responsible AI moves beyond policy documents and becomes an enforceable system capable of protecting citizens, workers and communities as AI use expands across government, business and daily life.

Source: Posttoday, global-index.ai