AISEAN

Global Policy Shift: Safety → Security

In February 2025, the UK rebranded its AI Safety Institute as the AI Security Institute, pivoting from ethics and bias concerns to national security threats — cyberattacks, bioweapons, and AI-enabled crime. The move signals a global governance shift that directly impacts ASEAN, where most national AI frameworks still focus on economic opportunity rather than security risk. aisi.gov.uk ↗

▾

AI dramatically lowers the cost and skill threshold for sophisticated cyberattacks. The UK AISI identified AI-enhanced cyber threats as among the most likely to manifest with high impact in the near term. For ASEAN, this is particularly acute: the region's rapid digitization has outpaced its cybersecurity infrastructure.

ASEAN exposure: Singapore's CSA and Thailand's NCSA have begun AI-specific cyber threat assessments. Indonesia, Philippines, and Vietnam remain significantly under-resourced. The 2024 ASEAN Cybersecurity Cooperation Strategy predates AI-enhanced attack vectors.

▾

The UK AISI detected multiple serious vulnerabilities in frontier AI models that could enable bioweapon development — these were fixed before model release. AI's ability to provide "uplift" to bad actors seeking to develop biological weapons is a top-tier concern for the AISI and the US CAISI (formerly NIST AI RMF).

ASEAN exposure: No ASEAN member state currently has a dedicated AI-biosecurity assessment framework. Vietnam and Indonesia have active biotech sectors that intersect with AI research without biosecurity guardrails specific to AI. Singapore's DSO is the only regional body with capacity approaching this risk domain.

▾

LLM-generated synthetic media and targeted political messaging have been deployed in elections across the Philippines (2025), Indonesia (2024), and Thailand (2025). The AISI's research on "levers of political persuasion with conversational AI" (published in Science, 2025) found LLMs can measurably shift political opinions.

ASEAN exposure: ASEAN is among the world's highest-risk regions: high social media penetration, contested elections, weak platform accountability, and multiple active geopolitical influence operations (South China Sea, Myanmar junta). The Philippines' DICT and Indonesia's Kominfo have issued warnings but lack enforcement capacity.

▾

The UK AISI established a dedicated Criminal Misuse Team in 2025, focusing on AI-enabled fraud, financial crime, and child sexual abuse material (CSAM) generation. Southeast Asia is a global epicenter for AI-enabled financial scams — the Myanmar and Cambodia cyber-scam compounds have increasingly adopted AI voice cloning and deepfake technology.

ASEAN exposure: The ASEAN 2024 Digital Ministers' Meeting flagged online scams as a top regional priority. Singapore's SPF and MAS have deployed AI-fraud detection. Myanmar's military-controlled scam operations represent a state-level threat vector unique to the region.

▾

The 2025 International AI Safety Report (led by Yoshua Bengio) highlighted that risks may come not only from malicious actors but from AI models themselves behaving autonomously in unintended ways. As ASEAN governments and militaries increasingly deploy AI in critical infrastructure and decision-making, the absence of human oversight frameworks is a structural gap.

ASEAN exposure: Vietnam's AI Law (2025) and Singapore's Model AI Governance Framework include human oversight provisions, but implementation guidance for high-stakes domains (military, critical infrastructure, judicial) is thin or absent across the region. No ASEAN member has ratified a binding human-control standard.

🗄️

We maintain and freely publish a continuously updated database of AI companies, funding rounds, open source projects, training datasets, LLM models, GPU infrastructure, and job listings across all 10 ASEAN member states.

All data is downloadable in CSV format at no cost. The database is updated monthly and covers 447 companies, 104 verified funding rounds, 139 open source projects, and 47 training datasets as of 2026.

📄

We produce short-form policy briefs (4–8 pages) analyzing AI governance gaps, sovereignty risks, labor displacement trends, and regulatory developments across Southeast Asia. All briefs are published under Creative Commons license and freely downloadable.

Target audiences include ASEAN national AI offices, academic researchers, journalists, and civil society organizations who cannot access expensive analyst reports.

🌱

We actively surface AI companies, researchers, open source projects, and datasets from lower-visibility ASEAN countries — Vietnam, Cambodia, Myanmar, Laos, and Brunei — that are systematically underrepresented in global AI coverage dominated by Singapore and Indonesia.

This includes a community submission system, monthly curation of new entries, and direct outreach to national university AI labs and government AI offices.

Dat Nguyen

Founder & Executive Director

CEO of Black Gibbon AI. Vietnamese-American, dual US/Vietnamese citizenship. 10+ years operating AI data infrastructure businesses across Southeast Asia.

[Board Member — Recruiting]

AI Policy / Academic

Seeking: researcher at an ASEAN university AI lab or think tank. Target institutions: NUS, NTU, AIT Bangkok, UMalaya, ITB Bandung, VinUniversity.

[Board Member — Recruiting]

Civil Society / Digital Rights

Seeking: practitioner from an ASEAN digital rights or civil society organization. Target orgs: EngageMedia, Internews, Digital Defenders Partnership, AccessNow SEA.

[Board Member — Recruiting]

Industry / Operator

Seeking: founder or senior operator at an ASEAN AI company. Ideally from Vietnam, Indonesia, Malaysia, or Philippines to ensure geographic diversity on the board.

~6 pages · Estimated audience: AI policy offices, academic researchers, NGOs

~8 pages · Estimated audience: Vietnamese government, labor NGOs, development finance institutions

~6 pages · Estimated audience: development finance institutions, impact investors, ASEAN national AI offices