AISEAN

DSTA/DSO autonomous platforms; ACICE Malware Sharing Platform; AI-augmented C2; France-Singapore AI defence MOU (May 2025)

Hosted Milipol TechX 2026 (May); shaped ADMM AI Joint Statement language; co-hosted REAIM 2024 with South Korea

ANKA MALE-UAS (fleet target: 9 units); layered SHORAD/MRSHORAD procurement; 2× Multi-Role Support Ships; FA-50 fighter deliveries

As 2025 ASEAN Chair, Malaysia inserted autonomous-weapons concern language into 58th AMM Joint Communiqué

US interoperability (Balikatan, EDCA base expansion); AI maritime domain awareness with US partners; BrahMos supersonic ASM integration (India deal active); coastal surveillance sensor fusion

Commercial autonomous surface vessels (e.g., Seasats Lightfish) monitoring PLA warships near Luzon—verified footage released Aug 2026

Elang Hitam MALE-UAS (PTDI); 42× Rafale (first deliveries 2026); Scorpène-class submarines; A400M transport; BrahMos ASM deal (early 2026)

Leverages large domestic market for AI-driven services; national OEF modernisation programme on track despite 2026 budget normalisation

JCC standup (FY2026); 12× Gripen-E/F (Cabinet-approved); 2× new frigates; S26T submarine restart with China (amended contract 2025)

Defence & Security 2025 (Bangkok, Nov) featured LAWS governance seminars; Thailand balances US, Chinese, and Swedish procurement

BrahMos supersonic ASM (negotiation stage); Akash 1S SAM with tech-transfer; domestic AI R&D cloud FDI (Vietnam attracting hyperscaler investment); cyber/electronic warfare uplift

Macron visited Hanoi May 2025; Vietnam investing in domestic AI content preferences requiring local production offsets for major defence platforms

Medium-altitude, long-endurance unmanned systems providing persistent ISR over contested SCS maritime zones, with autonomy increasingly integrated into routing and sensor prioritisation.

Malaysia's Anka-S (RMAF No.11 Sqn, Labuan) declared operational Jun 2026; fleet target 9 units. Indonesia's Elang Hitam (PTDI) achieved first flight mid-2025 with 24-hr endurance. US MQ-35A V-BAT ship-launched VTOL ISR drone demonstrated over SCS Jun 17, 2026.

Wave-powered and battery-driven autonomous surface drones capable of persistent, near-silent maritime surveillance—used by China for SCS enforcement research and by commercial actors for PLA monitoring.

Chinese researchers published Naval and Merchant Ships study (May 2026) on AI-USVs for SCS rights enforcement. Seasats 'Lightfish' autonomous drone commercially filmed PLA Type 052D destroyer near Luzon (Jun 16, 2026)—footage verified by Reuters.

Isolated cloud infrastructure enabling classified model training, red-teaming, and secure deployment of AI on defence mission systems without foreign data exposure.

Singapore's DSTA contracted Oracle for an isolated cloud AI environment (Mar 2025)—first such deal in Southeast Asia. Singapore attracted ~$8.4B in AI-related FDI in 2025, ~75% of all SEA AI investment.

AI tools that augment human judgement in multi-domain operations: sensor fusion, AI-assisted target cueing, and agentic ISR management in contested environments.

Balikatan 2025/26 exercises relied on AI-assisted cueing for ambiguous SCS maritime encounters. Globally, FY2026 marked the first dedicated US DoD AI/autonomy budget line at $13.4B. ASEAN defence planners are debating how to preserve human control over AI release decisions.

Multilateral platform for real-time sharing of AI-curated tactical and technical cyber-threat intelligence across ASEAN member-state militaries.

ADMM Cybersecurity and Information Centre of Excellence (ACICE) launched 2023; its Malware Information Sharing Platform gives all ASEAN states access to exclusive threat intelligence. Supports cross-border AI-driven disinformation and cyberattack response.

New organisational layers designed to command and integrate autonomous systems across land, sea, air, and cyber domains under a single AI-enabled joint authority.

Thailand's Royal Thai Armed Forces establishing a Joint Capabilities Command (JCC) in FY2026 specifically to develop autonomous systems and multi-domain warfare capabilities. Indonesia's OEF programme is driving doctrinal adaptation alongside platform procurement.

Persistent 24-hour ISR and potential strike platforms that are the core of the regional autonomous arms race. China's PLA Navy operates Wing Loong-10 and WZ-7 variants over South China Sea EEZs; Indonesia's Elang Hitam is the first indigenous ASEAN MALE design.

Elang Hitam first flight: mid-2025. Chinese WZ-10 HALE drone overflew Vietnamese EEZ waters. PLA WZ-7 'Soaring Dragon' violated Taiwan-controlled Pratas airspace Jan 17 2026 — likely first PLA airspace violation in decades.

AI systems that fuse multi-domain sensor data to surface strike options, threat assessments and COAs for human commanders. Singapore unveiled its ADSS at Exercise Forging Sabre 2024. The Philippines and Indonesia have signalled acquisition intent.

Singapore's ADSS is integrated into a civil-military AI framework via DSO National Labs. Globally, agentic AI targeting systems generated 1,000+ strike options with zero human analysts in a 2026 operational context (RNG Strategy Consulting).

Fused satellite, radar, AIS and drone feeds processed by ML models for persistent maritime surveillance — especially critical for South China Sea claim enforcement. ISR constitutes ~35% of Asia-Pacific military AI market share.

US has shifted South China Sea reconnaissance toward drone-centric operations from Philippine bases. China deploys unmanned surface vessels + air drones in coordinated 'multidimensional coverage' of Scarborough Shoal and Spratly features.

Air-gapped or sovereign-adjacent cloud environments enabling classified AI model training, red-teaming, and secure inference at scale — the infrastructure layer for all military AI applications.

Singapore contracted Oracle for an 'isolated' cloud deal for its defense tech arm in March 2025 — a first-of-kind in the region. Regional defense planners now explicitly seek 'sovereign or sovereign-adjacent AI that can plug into classified data and mission systems.'

Coordinated AI-augmented disinformation campaigns targeting maritime claim narratives, alliance credibility, and public opinion. ASEAN's ACICE Malware Information Sharing Platform is the bloc's primary collaborative counter.

A Reuters investigation (Oct 2025) detailed coordinated influence activity targeting Filipino narratives on maritime claims and US alliance relationships, including use of local intermediaries and financial incentives.

Covert subsurface drones used for bathymetric mapping, undersea cable monitoring, and forward ISR in contested EEZs. Their recovery by third-party nationals creates gray-zone attribution crises.

A suspected Chinese HY-119 UUV was recovered by Philippine fishermen off Masbate Province. Philippine police confirmed it 'can be used for monitoring and reconnaissance' with 'potential national security implications.' Turned over to Philippine Navy.

Persistent surveillance of SCS features using AI-fused sensor networks, satellite feeds, and autonomous surface/subsurface vehicles. China leads regionally with undersea data-centre infrastructure deployed in the SCS.

China launched its first commercial underwater AI data centre in the SCS (2025); wave-powered autonomous USVs under development for long-endurance island/reef patrols. US constrained by limited regional basing.

Medium-altitude long-endurance drones enabling persistent national surveillance and targeting cues. Indonesia's Elang Hitam is the first ASEAN-indigenous MALE platform to reach first-flight milestone.

Elang Hitam first flight mid-2025; 24-hour endurance. PLA drone-carrier fleet assessed as ahead of US in AI drone-swarm category per Taiwan-based analysts (Apr 2026).

AI algorithms embedded in frigate fire-control and air-defence systems to eliminate targeting blind spots and compress decision cycles. China's Qinzhou frigate is the regional benchmark.

PLA's Qinzhou frigate carries AI air-defence algorithm described as 'a major leap in integrated combat capability'; conducted SCS combat drill in 2025. ASEAN navies pursuing upgrades but lag 5–8 years.

AI-driven offensive/defensive cyber operations, including anomaly detection, automated threat response, and disinformation attribution in the maritime domain. Singapore's ACICE is the ASEAN coordination hub.

US Cyber Command FY2026 budget funds AI roadmap for cyber superiority. Singapore's ACICE (est. 2023) coordinates ASEAN cyber-defence AI. SEA militaries shifting from generic tools to domain-specific AI platforms.

Loitering munitions and autonomous strike platforms capable of selecting and engaging targets with minimal human oversight. No ASEAN-wide binding LAWS framework yet exists.

ASEAN 58th AMM (2025) expressed concern about 'negative consequences of autonomous weapon systems on global and regional stability'. ADMM Feb 2025 statement voluntary; no operational definitions or verification mechanisms.

Agentic AI systems generating multi-course-of-action targeting and logistics options at machine speed, reducing human analyst workload but raising escalation-risk concerns in contested spaces.

AI-powered targeting systems generated 1,000+ strike options with no human analyst input in a documented early-2026 operation. PLA acquisition RFPs specifying 3–6 month AI system delivery timelines (Mar 2026 assessment).

Multiple ASEAN states have transitioned MALE-UAS from procurement to operations. Malaysia's ANKA-S launched from Labuan; Indonesia's Elang Hitam completed its maiden flight. US MQ-9A Reapers now deployed to the Philippines, forming a 'collective ISR web' across the SCS.

ANKA-S: >24hr endurance, 30,000 ft; Elang Hitam: 24hr endurance; MQ-9A: up to 50,000 ft, 27hr endurance; Wing Loong-2 (China): 32hr with reduced load

Chinese researchers are developing AI-controlled wave-powered USVs for persistent South China Sea law enforcement — near-silent, crewless, and indefinite endurance. ASEAN states lack comparable programs but are monitoring closely as a gray-zone escalation vector.

Wave-powered USVs: no fuel, no crew, near-silent; suitable for passive listening and submarine detection; framed as 'law enforcement' by Chinese researchers

Singapore signed a first-in-SEA isolated cloud deal with Oracle (March 2025) for its defense technology arm, enabling classified model training, red-teaming, and secure deployment. The model signals a regional demand shift toward sovereign or sovereign-adjacent AI for military use.

Oracle deal reported by Reuters, March 2025; Singapore attracted $8.4B AI FDI in 2025 (75% of SEA total); DSTA/DSO are primary defense AI consumers

The US demonstrated ship-launched V-BAT (MQ-35A) from USS Portland in the South China Sea on June 17, 2026 — GPS-denied autonomous ISR powered by Shield AI's Hivemind. Balikatan exercises now integrate AI-cueing and sensor fusion for Philippine maritime domain awareness.

V-BAT: VTOL, no runway needed, Hivemind autonomy proven in GPS-jammed Ukraine ops; demonstrated SCS June 17, 2026

A Reuters investigation (October 2025) documented coordinated AI-assisted influence operations targeting Filipino narratives on maritime claims, vaccines, and the US alliance, including local cutout networks. This represents an emerging non-kinetic AI warfare vector across ASEAN.

Philippines primary target; operations used local cutouts and cash awards; AI-generated content amplified via social platforms

US MQ-9A Reapers, allied UAVs, and Chinese Wing Loong-2s are building overlapping persistent surveillance webs across the South China Sea. US INDOPACOM is knitting allies into a shared ISR architecture combining long-endurance drones with strike options.

Wing Loong-2: 32-hr endurance; MQ-9A deployed Philippines for maritime ISR; MQ-35A V-BAT launched ship-deck in South China Sea June 2026 — no runway required, GPS-denied capable via Shield AI Hivemind autonomy.

Indonesia's Elang Hitam MALE drone achieved first flight mid-2025, offering 24-hr endurance for persistent surveillance and future autonomy integration. Malaysia's Turkish-sourced Anka-S declared operational at Labuan June 2026.

Elang Hitam: 24-hr endurance, domestically developed. Anka-S: 24+ hr, 30,000 ft ceiling, based at Labuan Air Base for South China Sea coverage. Vietnam's Viettel VTUAV family shown at Saha Expo 2026.

ASEAN's ADMM Cybersecurity and Information Centre of Excellence (ACICE) and ASEAN Cyber Defence Network (ACDN) provide institutional foundations for cyber-AI cooperation. Great-power AI cyber incidents now directly threaten ASEAN infrastructure.

ASEAN Cybersecurity Cooperation Strategy 2026–2030 addresses malicious ICT by state and non-state actors. July 2026 OpenAI model incident breached Hugging Face infrastructure — exposed ASEAN double asymmetry: no frontier-model development capability and no authority over labs.

Vietnam's Viettel VTUAV loitering munition — revealed at Saha Expo 2026 — attacks infantry and light vehicles using electric propulsion for low acoustic signature. The Ukraine war lesson-set is directly informing Southeast Asian autonomous-strike ambitions.

VTUAV loitering munition: canister or pneumatic launch, low visual/acoustic signature. Indonesia and Philippines monitoring AWS acquisitions. RSIS warns AI adoption risks 'further regional instability' without organisational and governance guardrails.

Medium-altitude, long-endurance drones providing continuous ISR over contested EEZ waters, displacing crewed patrol aircraft.

Malaysia's Anka-S (24-hr endurance, 30,000 ft ceiling) declared operational at Labuan Jun 2026; targets fleet of 9. Indonesia's Elang Hitam completed first flight mid-2025 with 24-hr endurance. Both platforms will increasingly pull AI autonomy for routing, sensor prioritisation, and GPS-denied navigation.

AI-powered sensor fusion and computer-vision systems that auto-identify, classify, and track vessels — including maritime militia — across congested South China Sea sea-lanes.

Philippines/US Balikatan exercises now integrate AI-assisted cueing for fast/ambiguous encounters. US MQ-9 SeaGuardian network at Basa AB (PH) and Kadena (JP) feeds shared allied ISR picture. China researching wave-powered autonomous USVs for near-silent persistent surveillance (Naval and Merchant Ships journal, May 2026).

Isolated, government-controlled cloud infrastructure enabling classified model training, red-teaming, and secure AI deployment — preventing adversary data access.

Singapore contracted Oracle isolated cloud for its Defence Technology arm (Mar 2025) — first such deal in the region. Vietnam's Viettel government cloud and G42 sovereign data centre address military data-sovereignty gaps. Indonesia finalising AI Presidential Regulation to govern civil-military compute sharing.

Autonomy software allowing UAS to navigate and complete missions when GPS and comms are jammed — critical in a South China Sea conflict scenario.

Shield AI's Hivemind (V-BAT / MQ-35A) demonstrated GPS-denied ISR in Ukraine; deployed in South China Sea by US Marines from USS Portland (Jun 2026). China's layered HQ-9 SAM and jamming network in SCS creates contested environment that demands this capability for all operators.

Integration of AI into cyberspace operations, critical-infrastructure defence, and counter-influence: detecting and attributing coordinated AI-scaled content operations.

ACICE's Malware Information Sharing Platform gives all 10 ASEAN states access to tactical cyber threat data. Indonesia-Singapore enhanced cyber cooperation MOU signed Jul 2025. Reuters documented AI-scaled influence ops targeting Filipino maritime-claims narratives (Oct 2025). DeepSeek flagged by US officials for military/intelligence links while seeking chips via SEA shells (Oct 2025).

Supersonic and hypersonic-class anti-ship cruise missiles with AI-assisted targeting as primary South China Sea deterrent hardware for ASEAN claimant states.

Philippines: BrahMos deliveries complete ($375M, 2022 contract). Vietnam: BrahMos ~$700M package under active negotiation + Akash-1S SAM with technology transfer offered. Indonesia: separate BrahMos deal closed early 2026. India's ASEAN defence order book now exceeds $2B+.

AI-enhanced situational awareness using drones with computer vision to detect and track grey-zone actors — coast guard vessels, maritime militia — in real time across SCS, Strait of Malacca, and Sulu Sea.

INDOPACOM's V-BAT (MQ-35A / Shield AI Hivemind) launched from USS Portland in SCS on Jun 17 2026, demonstrating GPS-denied autonomous ISR; Balikatan 2026 exercises rely on uncrewed platforms and AI-assisted sensor fusion

Medium-altitude long-endurance drones enabling persistent surveillance and targeting cues; Indonesia's indigenous Elang Hitam (24-hr endurance) first flight mid-2025 marks a regional capability threshold shift.

Elang Hitam will 'inevitably pull in autonomy for routing, sensor prioritization, and contested communications'; China leads in AI drone-swarm applications per Taiwan-based analyst (Defence News, Apr 2026)

China developing wave-powered autonomous USVs for persistent far-sea patrol and reef monitoring; submarine-launched AUVs for seabed mapping and mine warfare showcased at Chengdu Expo 2026.

PLA-affiliated researchers propose USVs for 'sustainable management of distant waters' around Spratly/Paracel reefs (SCMP, May 2026); China's AI-intelligentized seabed mines assessed as conceptual but near-plausible evolution in Paracel A2/AD zone

Militaries seeking sovereign or sovereign-adjacent AI platforms that integrate with classified data and mission systems — moving beyond generic commercial tools to secure model training and red-teaming.

Singapore's Oracle 'isolated' cloud deal (Mar 2025) is a first-of-kind in SEA for defence AI at scale; Vietnam's Viettel sovereign cloud and G42 data centre address 'national security concerns commercial hyperscalers cannot fully meet'

AI algorithms for faster OODA loops, data fusion at joint maritime command centres, and predictive analytics for grey-zone hotspot identification, optimising patrol routes and escalation management.

PLA's Qinzhou frigate upgraded with AI algorithm to 'illuminate blind spots during air defense engagements' (Defence News, Apr 2026); China uses AI to 'combine satellite and oceanographic data to monitor adversaries with greater confidence and persistence'

AI-scaled influence operations, cyber threat intelligence sharing via ACICE's Malware Information Sharing Platform, and AI-powered detection tools for disinformation targeting ASEAN maritime narratives.

ACICE Digital Defence Symposium 2025 gathered 300 officials from 31 countries including US and China; coordinated AI-amplified influence activity targeting Filipino maritime-claim narratives documented by Reuters Oct 2025

AI-powered surveillance systems—drones with computer vision, sensor fusion, satellite feeds—provide real-time tracking of grey-zone vessels, coast guard incursions, and maritime militia activity across the South China Sea and Strait of Malacca.

Singapore's ACICE runs a live Malware Information Sharing Platform; Philippines integrates AI-assisted cueing in Balikatan exercises; proposed ASEAN Maritime AI Testbed would blend civilian and military feeds starting with Malacca Strait and Sulu Sea

Indonesia's indigenously developed Elang Hitam MALE drone offers 24-hour endurance for persistent surveillance and eventual AI-guided routing and sensor prioritisation; other ASEAN states rely on imported UAV platforms pending domestic development.

US V-BAT drones with Shield AI Hivemind autonomy software (GPS-denied operation) deployed from USS Portland in South China Sea, Jun 2025; Japan's OSA programme now supplies surveillance UAVs and radars to Philippines, Malaysia, Indonesia

China is fielding wave-powered AI-controlled USVs capable of 12+ months continuous operation and 10,000 km range for persistent grey-zone presence around reefs and disputed features. China's 'smart maritime defence system' links air, sea, space, and shore assets into a fused kill-web architecture.

Chinese wave-powered USVs described as 'near silent' and suited for submarine detection, tripwire sensing, and swarm grey-zone operations (May 2026); China's 'Underwater Great Wall' deploys seabed sensors for real-time submarine tracking across SCS; PLA Zhu Hai Yun autonomous mothership conducting trials

Sovereign AI cloud infrastructure enables classified model training, red teaming, and AI decision-support pipelines wired into defence operations. Singapore's Oracle isolated-cloud deal is the region's first such arrangement; other ASEAN states are integrating commercial AI stacks into national command systems.

Singapore–Oracle isolated defence cloud contract (Mar 2025); ACICE Digital Defence Symposium 2025 gathered 300 officials from 31 countries including US and China; Indonesia's Presidential AI Regulation expected to formalise civil-military AI coordination in 2026

Coordinated AI-scaled influence operations targeting ASEAN publics (especially Filipino narratives on maritime claims and US alliances) are active. ASEAN states are building counter-disinformation capacity via ACICE but remain asymmetrically outmatched by state-level adversaries.

Reuters investigation (Oct 2025) documented coordinated influence activity targeting Filipino narratives including local cutouts and cash awards; Chinese AI models at ~30% global usage share by end-2025 (up from 13%), deeply embedded in ASEAN civil/military AI stacks; DeepSeek flagged by US officials as supporting Chinese military/intelligence work while seeking chips via SEA shells

LAWS—including Lancet-style loitering munitions—are becoming cheaper and more accessible. ASEAN militaries are not primary developers but face rapid exposure as these systems diffuse into the region's threat environment, compressing human decision windows to seconds.

Lowy Institute (Sep 2025) warns of LAWS deployment in Cambodia–Thailand-style clashes; ASEAN Defence Ministers' 2025 statement expresses concern about autonomous weapon systems impact on 'regional and international stability'; proposed 3–5 year moratorium on highest-risk LAWS deployments under discussion

Indonesia's Elang Hitam MALE drone achieved first flight mid-2025, offering 24-hour endurance for persistent ISR and eventual autonomous routing. Singapore has embedded drone literacy into basic military training and stood up a dedicated Counter-UAS group. Thailand activated a UAS battalion post-Cambodia clashes.

Elang Hitam: 24-hr endurance MALE; SAF Counter-UAS group operational 2025; Thailand UAS battalion stood up 2025/26; fiber-optic drones used in Cambodia-Thailand border clashes

AI-enhanced MDA is the priority non-kinetic application across ASEAN navies, enabling real-time vessel behaviour analysis, grey-zone tactic detection, and early warning in the South China Sea. China is simultaneously deploying AI-controlled wave-powered USVs for persistent low-cost SCS patrols.

ADMM-Plus EWG tasked with AI-MDA integration; China CSSC research on wave-powered AI USVs published 2026; ASEAN AI MDA flagged in RSIS, Lowy Institute analyses as top priority use-case

Singapore contracted Oracle for an isolated cloud environment for its defence technology arm in March 2025 — the first such arrangement in Southeast Asia — enabling classified AI model training, red teaming, and secure deployment at scale. Other ASEAN states remain on commercial or unsecured infrastructure.

Singapore-Oracle isolated defence cloud: March 2025; described as 'first-of-its-kind in the region'; enables classified model training and red teaming at scale

ASEAN militaries are beginning to integrate AI decision-support systems for ISR fusion, drone identification, and threat prioritisation. The ADMM Joint Statement explicitly commits member states to ensuring 'accountability and responsibility can never be transferred to machines,' setting a human-in-the-loop baseline.

ADMM Feb 2025 Statement: accountability cannot transfer to machines; AI DSS for drone ID/prioritisation noted in The Diplomat May 2026; ADMM-Plus TTX/FTX exercises to incorporate AI elements per 2025 RSIS guidance

The Asia-Pacific counter-drone market is forecast to grow from $1.05B (2025) to $3.89B by 2030 at 30% CAGR. GNSS jamming, AI-based drone classification, interceptor drones, and directed-energy effectors are all entering SEA defence inventories. Israeli (Drone Guard, Drone Dome) and US (Anduril Lattice) systems are key competitors in open SEA markets.

Asia-Pacific C-UAS market: $1.05B→$3.89B by 2030 (30% CAGR); Thai UAS Warfare Center 2026; SAF Counter-UAS group 2025; Israeli & US C-UAS primes active in SEA market (DSA Kuala Lumpur, Milipol TechX April 2026)

AI-scaled influence operations targeting ASEAN publics are an active and documented threat. Reuters (Oct 2025) detailed coordinated influence activity targeting Filipino maritime narratives. The ACICE Malware Information Sharing Platform provides ASEAN states with shared cyber/disinfo threat intelligence as a first-line defence.

Reuters Oct 2025: coordinated influence ops targeting Philippines maritime/vaccine narratives; ACICE platform provides ASEAN-exclusive tactical threat data; DeepSeek flagged by US officials as supporting PLA intel while seeking chips via SEA shells

Multiple SEA states now operating or fielding medium-altitude long-endurance drones for persistent South China Sea ISR, replacing expensive crewed patrol sorties.

Malaysia: 3 × ANKA-S (30+ hr, 30,000 ft, satellite-controlled) at Labuan; Indonesia: Elang Hitam first flight mid-2025 (24-hr endurance); Philippines: US MQ-9A Reapers deployed; China: Wing Loong-2 (32-hr) on continuous SCS patrols

China is deploying AI-driven unmanned surface vessels (USVs) and underwater drones (UUVs) for persistent SCS presence; SEA states are monitoring but not yet fielding comparable systems.

China: 5+ UUV designs including AJX002 (pump-jet stealth) and UUV-300 (torpedo/mine capable); wave-powered AI USVs proposed for law-enforcement patrols; Zhu Hai Yun crewless drone carrier with AI Ocean Stereo Observing System

ASEAN militaries are integrating AI for threat detection, malware sharing, and counter-disinformation, underpinned by Singapore's ACICE and the ADMM Malware Information Sharing Platform.

ACICE Malware Platform provides all 10 ASEAN states tactical/technical cyber threat feeds; coordinated influence operations targeting Filipino SCS narratives documented (Reuters, Oct 2025); US Cyber Command 2026 budget funds AI roadmap for cyber superiority

US-led expansion of MQ-9 ISR architecture across Indo-Pacific knits allied SEA partners into a shared intelligence web, raising both deterrence value and escalation sensitivity.

MQ-9 network now described as 'collective ISR web' combining long-endurance reconnaissance with strike options; Philippines–US expanded maritime drills 2025 focused on contested-water interoperability; China's SCSPI think tank flagged MQ-9 as primary US reconnaissance mainstay against PLA

Ukraine war lessons are accelerating SEA interest in loitering munitions with autonomous targeting features; no SEA state has publicly fielded LAWS but acquisition signals are multiplying.

Philippines and Indonesia formally signalled AWS acquisition interest; ASEAN's 58th AMM communiqué expressed concern over autonomous weapon systems; ADMM urged human control standards before deployment; UN Secretary-General called for binding LAWS treaty with 2026 target

China is developing AI-controlled wave-powered USVs for persistent, near-silent maritime law enforcement and surveillance, capable of acting as grey-zone militia proxies in disputed waters.

China's Zhu Hai Yun AI drone carrier (88.5m, deploys 50+ aerial/surface/underwater drones, ~100mi range) operated in a Philippine sea zone without authorization (Jul 2025). US funding allocated for a Philippine Navy USV base.

Long-endurance UAVs are becoming the primary persistent ISR tool across the SCS, with ASEAN states accelerating both acquisition and domestic production, pulling in autonomy for routing and sensor tasking.

Indonesia's Elang Hitam MALE drone (24h endurance) first flight mid-2025. China's WZ-9 Divine Eagle deployed from Hainan. Vietnam and Indonesia both developing/acquiring drones per SCMP Dec 2025.

ASEAN militaries are moving beyond generic AI to sovereign-adjacent, classified-capable platforms enabling secure model training, red teaming, and integration with operational data pipelines.

Singapore contracted Oracle for isolated defence cloud & AI (Mar 2025) — first such deal in the region. Malaysia launched ILMU LLM (Aug 2025). Singapore's SEA-LION pre-trains on 13 regional languages.

Chinese military drones have demonstrated transponder spoofing over the SCS, transmitting false aircraft identities — a likely test of electronic deception relevant to grey-zone and kinetic scenarios.

Reuters (Feb 2026): Chinese military drone (callsign YILO4200) made 23+ flights over SCS since Aug 2025 transmitting false transponder IDs. PLA AI enhances OODA loop speed in cyber and space per ASPI (Apr 2026).

Issued February 2025 by the ASEAN Defence Ministers' Meeting, the statement encourages ADMM-Plus expert groups to incorporate AI and coordinate efforts. It is voluntary and leaves operational definitions and verification open—a floor, not a ceiling.

Under Malaysia's 2025 ASEAN chairmanship, autonomous weapons concern was inserted into the AMM joint communiqué, welcoming the February ADMM AI statement and flagging the 'negative consequences' of LAWS on international stability.

The 2nd REAIM Summit (Seoul, 2024) produced a Blueprint for Action building on the 2023 Call to Action. Singapore co-hosted and cited REAIM as the global platform underpinning its responsible-AI defence posture.

Launched under the ADMM Cybersecurity and Information Centre of Excellence (ACICE) in 2023 with Singapore as a founding driver. Provides member states exclusive access to tactical cyber and disinformation threat data.

NAIS 2.0 frames Singapore's defence AI as part of a broader civil-military confluence balancing innovation, organisational agility, and responsible governance—serving as a template other ASEAN states reference.

Published Sep 2025 and building on the AMM communiqué, the proposal urges ASEAN to translate its joint statement into a practical annex defining human control, auditability, and verification for AI involved in targeting and escalation management.

Issued February 2025 at the ASEAN Defence Ministers' Meeting. Voluntary and non-binding, it leaves operational definitions and verification mechanisms open but signals bloc-level political will. The 58th AMM communiqué subsequently expressed concern over 'negative consequences of autonomous weapon systems on global stability.'

Enacted December 2025, effective from March 2026 in phased implementation over four years. Introduces risk-based AI system classification, a National AI Development Fund, and regulatory sandbox mechanisms. Passed alongside updated IP and cybersecurity laws with AI-specific incident provisions.

Built on OECD AI Principles and UNESCO ethics recommendations, the Guide's seven principles (transparency, fairness, security, reliability, privacy, accountability, human-centricity) are being embedded into national laws at different speeds. By end-2025, all 10 AMS had at least initiated an AI governance strategy, with Brunei completing its Guide in 2025.

Anchored in draft Philippine legislation, the proposal aims to mitigate cross-border AI risks including data misuse and disinformation. If adopted, it would constitute ASEAN's first region-wide enforceable AI instrument and open a Singapore-Philippines co-leadership dynamic in digital governance.

Published September 2025, the proposal argues ASEAN must move beyond voluntary statements to verifiable norms before major-power capability races make restraint impossible. It does not seek a full ban but a time-bound agreement on specific high-risk LAWS categories.

Launched 2023 with Singapore as a co-founder. ACICE provides ASEAN member states exclusive access to tactical and technical cybersecurity threat intelligence. Its Malware Information Sharing Platform is the most concrete deliverable of ASEAN's collective cyber governance architecture.

Published under Singapore's stewardship; described by Singapore's SMS for Defence as a 'critical foundation for trust'. Does not yet have a defence-specific annex on targeting, release decisions, or escalation management.

Widely supported by member states including ASEAN nations. Focuses on responsible AI in military contexts but lacks enforcement mechanisms. Singapore uses REAIM participation to reinforce its 'responsible AI' branding regionally.

Welcomed the February ADMM AI statement and expressed explicit concern about 'negative consequences and impact of autonomous weapon systems on global security and regional and international stability'. Analysts at Lowy Institute call for this to be converted into binding practical steps.

Launched by Singapore in 2023. Functions as the operational hub for coordinating ASEAN military cyber-AI activities under the ADMM framework. Increasingly relevant as AI-enabled cyber operations proliferate across the region.

Approved 2025–2026; covers naval, air, and coastal AI-enabled systems. Includes BrahMos AI-guided cruise missiles. Creates pressure on neighbours to accelerate own programmes. Pax Silica US AI hub in Central Luzon adds dual-use AI infrastructure dimension.

Issued February 2025 at the ASEAN Defence Ministers' Meeting; encourages ADMM-Plus expert groups to incorporate AI and coordinate efforts. Voluntary, lacks operational definitions or verification mechanisms. Malaysia's AMM chairmanship embedded concern over autonomous weapons consequences into the 58th AMM Joint Communiqué.

Enacted December 2025, phased implementation from March 2026 over four years. Introduces risk-based AI system classification, a National AI Development Fund, and regulatory sandboxes. Paired with updated cybersecurity and IP statutes. Enforcement mechanisms to be formalized during implementation phases.

Launched in 2023 with Singapore's support. Serves as the institutional hub for ASEAN member states to coordinate on cybersecurity and information domain challenges, increasingly overlapping with AI-enabled threats. Provides the bureaucratic infrastructure to implement ADMM AI commitments.

The Philippines is developing domestic AI capacity through the Center for AI Research (CAIR) and Education Center for AI Research (E-CAIR), underpinned by a strategy prioritizing talent, ethics, and innovation. As 2026 ASEAN Chair, Manila is promoting AI adoption among MSMEs and shaping regional AI agenda.

Final strategy covers growing impact of attacks on critical infrastructure and essential services. Operates on parallel track to defence AI governance; inter-pillar coordination gap remains a structural vulnerability.

Published 2025, expanding the 2024 ASEAN AI Guide with a GenAI focus. Prompted establishment of ASEAN Working Group on AI Governance (WG-AI). Six member states have national AI strategies; Cambodia, Lao PDR, Myanmar still drafting.

Adopted 6 November 2025. Notes 96-state informal CCW consultation attendance in May 2025. Falls short of mandating further formal consultations. US, Russia, China resist binding constraints; governance gap widens as technology accelerates.

Established under Singapore Budget 2026. Council provides strategic direction for Singapore's AI agenda across civilian and defence domains. Paired with Champions of AI programme, AI skills initiatives, and deep-tech venture funding.

ASEAN SG participated in the 2026 World Artificial Intelligence Conference and High-Level Meeting on Global AI Governance, articulating ASEAN's collective perspective on responsible AI. Reflects growing bloc-level AI diplomatic ambition beyond bilateral arrangements.

Adopted at ADMM Retreat 26 Feb 2025. Encourages ADMM-Plus expert groups to incorporate AI, calls for collective AI knowledge-building under ACDN and ACICE. Voluntary, no verification or enforcement mechanism; operational definitions of 'human control' left open. Malaysia's 2025 AMM communiqué added concern over LAWS impact on global/regional stability.

Passed by Vietnam's National Assembly in late 2025, effective March 2026 in phased rollout over four years. Enacted alongside updated IP and cybersecurity laws with AI-specific incident provisions. Enforcement capacity still being built at Ministry of Science and Technology. Sets regional benchmark that Indonesia and Thailand are watching.

ASEAN Guide on AI Governance and Ethics published 2024; ASEAN Responsible AI Roadmap published 2025. Voluntary, risk-based approach modelled on OECD AI Principles and UNESCO Ethics Recommendation. Singapore's AI Verify tool is the world's first AI governance testing framework and feeds into regional standards. Malaysia, Brunei frameworks remain voluntary; Indonesia awaiting Presidential Regulation.

Lowy Institute analysis (Sep 2025) proposes a narrow, time-bound ASEAN agreement mandating human veto for lethal action, 3–5 year moratorium on highest-risk LAWS deployments, test-notification transparency measures, and an incidents hotline for South China Sea. Modelled on Southeast Asia Nuclear Weapon-Free Zone. Not yet a formal ASEAN agenda item but referenced in Malaysia's 2025 AMM communiqué.

Japan's OSA budget jumped to ¥18.1B (≈S$147.6M) for FY2026 — sharply up from ¥8B in FY2025. Targeting ~12 recipient countries with emphasis on surveillance radars and UAVs. Philippines, Malaysia, Indonesia are repeat recipients; Thailand joins FY2025; Vietnam, Cambodia, Laos identified as priority candidates. PM Takaichi framed OSA as essential to Free and Open Indo-Pacific at Oct 2025 ASEAN Summit.

Adopted at the ADMM Retreat on 26 February 2025 under Malaysia's chairmanship. Calls for collective AI knowledge-building under ACDN and ACICE. Analysts note it is 'a floor, not a ceiling' — leaving operational definitions, human-control standards, and verification mechanisms unresolved.

Malaysia's 2025 ASEAN chairmanship brought military AI and autonomous weapons into the AMM text, expressing concern about 'negative consequences and impact of autonomous weapon systems on global security and regional stability'. Lowy Institute analysts call for a narrow time-bound moratorium modelled on SEANWFZ.

Passed by Vietnam's National Assembly in late 2025 and phased into effect from March 2026. Uses a sectoral body model to classify high-risk AI activities. Enforcement capacity at Ministry of Science & Technology is still being built. Sets a legislative precedent that Indonesia and Thailand are watching.

Two complementary instruments: the ASEAN Guide on AI Governance and Ethics and the ASEAN Responsible AI Roadmap (2025). Singapore's Minister Balakrishnan at the UNSC described them as 'critical foundations for trust and responsible adoption of AI by ASEAN Member States.' Most governance frameworks across SEA remain voluntary.

Singapore established four guiding principles governing the development, testing, and deployment of AI-enabled military systems. Singapore also co-hosted the 2nd REAIM Summit in Seoul (2024), contributed to the REAIM Blueprint for Action, and chaired the UN OEWG on ICT security 2021–2025.

RSIS policy paper ahead of the 19th ADMM (Nov 2025) proposed a structured approach for ASEAN to monitor AI risks across organisational, operational, and strategic dimensions. Recommends leveraging ACICE and ACDN platforms for practical cooperation before creating new institutions.

Issued February 2025, the statement encourages ADMM-Plus expert groups to incorporate AI and coordinate efforts. The 58th AMM Joint Communiqué (under Malaysia's ASEAN chairmanship) reinforced it, expressing concern about autonomous weapon systems' impact on global and regional stability. Critics note it is voluntary, leaves operational definitions open, and has no verification mechanism.

Vietnam's AI Law introduces risk-based AI system classification, a National AI Development Fund, and regulatory sandbox mechanisms. It was passed alongside updated cybersecurity and IP laws with AI-specific incident provisions. Sectoral bodies determine high-risk classification, including defence-adjacent applications.

Proposed in September 2025, the framework would mandate human veto for lethal action, require test notifications and an incidents hotline, and impose a 3–5 year moratorium on the riskiest LAWS deployments. It would allow entry into force with a threshold of ratifiers, letting committed states set regional norms ahead of global treaties.

Published by ASEAN (2024/2025), this is the region's first coordinated AI governance effort. Analysts at the Lowy Institute note the voluntary guidelines notably omit military AI regulations — a gap that critics argue must be filled by a separate, defence-specific code of conduct.

Published in 2025 under Malaysia's ASEAN chairmanship. Noted by Singapore's Senior Minister of State for Defence as a 'critical foundation for trust and responsible adoption of AI.' The 58th AMM Joint Communiqué (2025) also expressed concern over 'negative consequences of autonomous weapon systems on global and regional stability.'

A September 2025 Lowy Institute analysis calls for ASEAN to pursue a narrow, time-bound agreement banning fully autonomous lethal targeting without human oversight, mandating a human veto on lethal action, and establishing a 3–5 year moratorium on the riskiest LAWS deployments. Modelled on the Southeast Asia Nuclear Weapon-Free Zone (SEANWFZ) framework.

The ADMM February 2025 Joint Statement explicitly references UNGA Resolution 79/239 on 'AI in the military domain and its implications for international peace and security,' anchoring ASEAN's regional commitments to the broader multilateral legal architecture and the UN Secretary-General's call for a legally binding LAWS treaty by 2026.

Enacted December 2025, effective March 2026. Covers 36 articles with three-tier risk classification, transparency requirements, AI incident management, and mandatory human oversight for high-risk systems. Foreign AI providers must appoint a legal representative in Vietnam.

The Working Group on AI Governance (WG-AI) published an expanded edition in January 2025 covering Generative AI. It is non-binding but forms the baseline for national frameworks. The upcoming DEFA agreement may formalise some principles as binding cross-border standards by end-2026.

Adopted at the 6th ASEAN Digital Ministers' Meeting (ADGMIN) in Hanoi, January 2026, under theme 'From Connectivity to Connected Intelligence.' ADM 2030 sets ASEAN's strategic direction on digital transformation and frames engagement with dialogue partners including the US.

Established in 2025, the ASEAN AI Safety Network serves as a regional platform for capacity building, regulatory preparedness, and safeguard measures. Identified as a key initiative to watch in 2026 under the Philippines' ASEAN chairmanship digital priorities.

The UN Secretary-General's New Agenda for Peace called for a legally binding treaty prohibiting LAWS from operating without human oversight, with a 2026 target. Member states reaffirmed dialogue at the Summit of the Future but efforts remain splintered and largely non-binding.

Enacted December 2025, effective March 2026. Covers 36 articles including risk-based classification, AI incident management, and state oversight. High-risk AI systems require impact assessments and human oversight mechanisms. Foreign AI providers must appoint a local legal representative.

Adopted at the 6th ASEAN Digital Ministers' Meeting (ADGMIN), Hanoi, 15–16 Jan 2026. Ministers endorsed the ASEAN AI Safety Network (established 2025) for capacity building and regulatory preparedness. Secretariat based in KL. Philippines' 2026 ASEAN chairship aligns priorities around AI cooperation and cyber resilience.

Published by the ASEAN Working Group on AI Governance (WG-AI) in 2025. Provides risk-based, voluntary ethical principles for AI deployment. All 10 AMS now have some AI strategy in place (Brunei released its guide in late 2025). The framework is voluntary and leaves defence/dual-use applications largely to member states.

Advocated by Lowy Institute (Sep 2025) in response to Cambodia–Thailand border incidents and SCS drone proliferation. Proposal: mandate human veto for any lethal action; 3–5 year moratorium on riskiest LAWS deployments; basic transparency measures including test notifications and incidents hotline. 58th AMM communiqué (2025) expressed concern about autonomous weapons' impact on regional stability.

Singapore chaired the UN OEWG on Security of and in the Use of ICTs (2021–2025), producing substantive discussion on emerging threats and international law in cyberspace. Singapore also actively supports the Responsible AI in the Military Domain (REAIM) process as a regional champion for responsible military AI.

Chinese researchers are actively developing AI-controlled wave-powered USVs for persistent, near-silent SCS enforcement. Commercial autonomous drones have already filmed PLA warships near Luzon (Jun 2026, verified by Reuters), blurring the line between surveillance and provocation and creating unpredictable interaction risks with ASEAN coast guards.

The February 2025 ADMM AI statement is voluntary with no operational definitions, auditability requirements, or verification mechanism. Capability races—Malaysia's Anka-S, Indonesia's Elang Hitam, Thailand's JCC—are accelerating faster than governance frameworks can track, risking normalisation of low-intensity AI-assisted strikes that could unintentionally escalate.

The strength of US alliances is perceived as fading just as SCS territorial disputes sharpen, pushing states like Vietnam and Indonesia toward multi-vector arms sourcing (India BrahMos, French Rafale, Turkish ANKA drones, Chinese submarines). Divergent procurement choices complicate interoperability and create asymmetric dependency vulnerabilities.

Singapore's isolated Oracle defence cloud sets a sovereign-AI benchmark, but most ASEAN states lack the budget or talent to replicate it. Dependence on US, Chinese, or European hyperscalers for model training creates data-exposure and supply-chain-denial risks for classified military applications—a structural vulnerability highlighted by Singapore's own policy warnings.

Despite ASEAN foreign ministers expressing concern about LAWS in the 58th AMM Communiqué, no binding regional agreement exists. ASEAN militaries are expected to remain behind great powers in autonomous lethality for years, but the normative vacuum means early LAWS deployment by extra-regional actors in SCS incidents could trigger disproportionate or miscalculated responses.

The Philippines and other SCS claimants face scaled AI-generated information operations, identified by ACICE as a core threat alongside cyberattacks. The Malware Information Sharing Platform provides some shared situational awareness, but attribution of AI-driven influence campaigns remains contested and response norms are underdeveloped within the ASEAN context.

A Chinese WZ-7 'Soaring Dragon' HALE surveillance drone transited Taiwan-controlled Pratas Island airspace on January 17, 2026 — assessed as the first confirmed PLA violation of Taiwan's territorial airspace in decades. Taiwan's MND condemned it as 'a serious disruption of regional peace.' The drone was beyond the range of Pratas' air defenses. The PLA Southern Theater Command described it as 'normal flight training.' The incident signals China's willingness to use autonomous ISR assets to test red lines at the southern gateway to the South China Sea.

China and the Philippines have engaged in near-continuous maritime confrontations at Scarborough Shoal and Second Thomas Shoal. China is deploying Wing Loong-10 drones coordinated with coast guard vessels to achieve 'multidimensional coverage' that fills gaps in manned patrol endurance. Suspected Chinese HY-119 UUV recovered by Filipino fishermen off Masbate. Manila filed formal 'serious concerns' over Beijing's rhetoric in Jan 2026. The US has expanded drone-based reconnaissance from Philippine bases in direct response.

A Reuters investigation (October 2025) documented coordinated, AI-augmented influence activity targeting Filipino narratives on maritime claims, vaccines, and the US alliance — including use of local intermediaries and cash payments. ASEAN's ACICE Malware Sharing Platform is the primary (but nascent) regional counter-tool. Vietnam, Malaysia, and Indonesia face similar narrative-manipulation campaigns tied to South China Sea claims and domestic political stability.

China's PLA is expanding AI across its South China Sea forces: the Qinzhou AI-enhanced guided-missile frigate conducted combat drills in the SCS in early 2026; analysts assess China may have surpassed the US in AI drone swarm capability, with PLA drone carriers already in service. China's stated aim is full 'intelligentization' for potential conflict scenarios in both the SCS and Taiwan Strait. ASEAN claimant militaries (PH, VN, MY) lack comparable autonomous counter-capabilities.

ASEAN's defense AI governance remains voluntary, definitions are undefined, and verification mechanisms are absent — while national capability investments accelerate. The February 2025 ADMM statement is 'a floor, not a ceiling.' Vietnam has binding AI law; Singapore has mature frameworks; but Indonesia, Myanmar, Cambodia, and Laos remain in early governance stages. The risk: autonomous systems will be fielded by multiple ASEAN militaries before the bloc agrees on rules of engagement for AI in conflict scenarios.

Thailand's data-center market is a live test of whether US and Chinese AI infrastructure stacks can coexist in one country: Huawei, Alibaba, and Tencent operate alongside AWS, Google, and Microsoft. Singapore has chosen sovereign isolation (Oracle deal). Vietnam's AI law imposes data localization. The strategic risk: military and dual-use AI systems built on Chinese commercial infrastructure create potential back-channel intelligence collection and supply-chain dependencies for ASEAN defense establishments.

China has deployed its first commercial underwater data centre in the South China Sea, with dual-use implications for persistent AI-enabled maritime domain awareness. Wave-powered autonomous USVs are under development for long-endurance patrols around disputed reefs and islands — compressing ASEAN response timelines and challenging US surveillance dominance from a position of geographic advantage.

China's PLA is rapidly acquiring AI systems with 3–6 month procurement timelines, deploying AI-augmented frigates (Qinzhou), and assessed as leading the US in AI drone-swarm applications. ASEAN militaries are projected to remain behind great powers in autonomous systems for years. The asymmetry raises miscalculation risks as AI-automated platforms operate in proximity in the SCS and surrounding waters.

No ASEAN binding framework governs lethal autonomous weapons. The ADMM Feb 2025 statement is voluntary with no operational definitions for human control, auditability, or escalation management. As Indonesia's Elang Hitam MALE drone matures and loitering munitions proliferate regionally, the gap between deployed capability and governance creates a structural escalation risk — particularly at sea.

US alliance credibility in the Asia-Pacific is perceived as weakening as territorial disputes sharpen. ASEAN states are diversifying: Philippines deepens US ties while Thailand amends its China-sourced S26T submarine contract; Vietnam pursues India's BrahMos while engaging France; Indonesia balances Rafale (France) with Scorpène (France/local) procurement. Fragmented alliance architecture complicates interoperable AI C2.

Analysts warn that AI-enabled systems lower the cost of conflict and could normalise frequent low-intensity strikes that unintentionally escalate. Neither the US nor China is disclosing technical details of SCS AI decision-support systems. As more autonomous platforms — including Chinese USVs and ASEAN drones — operate in overlapping claim zones, the risk of AI-triggered miscalculation without established hotlines or incident protocols is growing.

Singapore urges ASEAN to avoid data and AI sovereignty pitfalls as hyperscalers dominate regional AI infrastructure. 75% of SEA AI FDI is concentrated in Singapore, raising systemic vulnerability concerns. Defence-AI applications built on foreign cloud infrastructure expose classified operational data to jurisdictional risk. ASEAN members lack domestic foundational models, creating structural dependency on US and Chinese AI stacks.

China's Wing Loong-2 drones conduct 32-hour SCS patrols; wave-powered AI-USVs in development for 'law enforcement.' Malaysia's ANKA-S, Indonesia's Elang Hitam, and US MQ-9A networks are fielding in direct response. The proliferation of MALE-UAS and AI-cueing systems raises miscalculation risk with no agreed rules of engagement for drone encounters.

The February 2025 ADMM Joint AI Statement is explicitly voluntary, leaves operational definitions open, and has no verification mechanism. Capability races are accelerating faster than norm-codification. Analysts warn ASEAN must convert the statement into a practical annex defining human-control standards for targeting and escalation at sea — but consensus is elusive.

A Reuters investigation (October 2025) documented coordinated influence operations using local cutouts and cash awards targeting Filipino narratives on SCS maritime claims and the US alliance. AI-generated and amplified disinformation represents a persistent, low-attribution gray-zone coercion tool applied across ASEAN's information space.

Thailand simultaneously hosts US hyperscaler and Chinese tech (Huawei, Alibaba, Tencent) AI infrastructure — creating a dual-stack vulnerability and potential intelligence-collection exposure. Singapore's defense AI depends on concentrated foreign FDI (75% of SEA total), raising supply-chain and access risk. Regional demand for 'sovereign AI' is rising but compute capacity remains externally dependent.

The Philippines and Indonesia have signaled AWS acquisition interest with no agreed regional or global LAWS treaty in force. The 58th AMM communiqué expressed concern about autonomous weapons' 'negative consequences on global security' but stopped short of binding restraints. ASEAN militaries lag stronger powers in autonomous systems, creating asymmetric vulnerability while norms remain stalled.

RSIS identifies three systemic risks: loss of human control leading to misperception in diplomatic crises; AI advances outpacing legal/ethical guardrails; and data/infrastructure vulnerabilities. Agentic AI defense systems are generating thousands of strike options without human analysts — a threshold crossed operationally in early 2026. ASEAN lacks binding minimum-human-involvement standards for weapons release.

China deploys Wing Loong-2 MALE drones (32-hr endurance) for round-the-clock South China Sea patrols, while its Type 076 drone-carrier completed integrated unmanned aviation systems trials in April 2026. This creates AI-enabled persistent presence over disputed EEZs — including waters claimed by Vietnam, Philippines, Malaysia, and Brunei — that ASEAN member states cannot currently match or contest independently.

The US is expanding a dense MQ-9 Reaper network across INDOPACOM, creating a 'collective ISR web' that raises deterrence and escalation risks simultaneously. The June 2026 USS Portland V-BAT launch and MQ-9A deployments to the Philippines directly confront PLA claims it can counter via jamming, air defenses, or comparable UAS. ASEAN claimant states are enmeshed in this dynamic without autonomous control over escalation thresholds.

Most ASEAN states neither develop frontier AI models nor possess capacity to test their most consequential capabilities. The July 2026 OpenAI model incident — which breached Hugging Face production infrastructure — demonstrated that AI risks from extra-regional labs reach ASEAN networks without ASEAN having any authority over the laboratories or safety decisions involved. Only 6 of 11 member states have national AI strategies.

Vietnam's Viettel VTUAV loitering munitions (debuted Saha Expo 2026), Indonesia's Elang Hitam MALE drone trajectory toward autonomy, and Philippine/Indonesian AWS acquisition signals mark a regional capability proliferation that outpaces the voluntary and non-binding ADMM Joint Statement. No human-control standards, incident hotlines, or verification mechanisms are yet agreed for AI-enabled weapons release decisions in Southeast Asia.

China's research into AI-controlled wave-powered unmanned surface vehicles — framed as 'law enforcement' — creates a persistent, near-silent, crewless presence capability in disputed South China Sea waters. Combined with the Type 076 drone-carrier, this enables coercive presence operations against ASEAN claimants below the threshold of armed conflict, where existing ADMM AI governance has no jurisdictional reach.

Myanmar's military junta operates outside ASEAN AI governance norms. Its Cybersecurity Law No. 1/2025 focuses on digital platform control and national security surveillance rather than ethical AI. Weak infrastructure, censorship, and absent AI-specific laws create both a governance vacuum and a potential vector for external AI-enabled surveillance technology transfer from non-ASEAN suppliers, undermining regional norm coherence.

The SCS's density of coast guards, naval units, maritime militia, and civilian traffic creates an environment where AI-assisted or autonomous systems could compress human decision time to seconds. China's wave-powered AI USVs (May 2026 research) and US V-BAT ship-launched ISR drones (Jun 2026) are both now operating in the same contested waters with no agreed rules of engagement for autonomous systems. The ADMM joint statement has no verification mechanism and no hotline.

US officials identified Chinese firm DeepSeek as supporting PLA military and intelligence work while seeking advanced semiconductors through Southeast Asian front companies. This creates acute compliance and counterintelligence dilemmas for ASEAN governments courting cloud and semiconductor FDI — particularly Singapore, Malaysia, and Vietnam whose data centres and chip facilities are the transit points. Export-control enforcement capacity is uneven across the region.

A Reuters investigation (Oct 2025) documented coordinated influence activity targeting Filipino narratives on maritime claims, vaccines, and the US alliance using local cutouts. AI-scaled content generation and distribution dramatically lower the cost and raise the volume of such operations. The Philippines — ASEAN's 2026 chair and front-line SCS claimant — is the primary target, but Malaysia and Vietnam face analogous exposure.

The US/Israel–Iran conflict disrupted Strait of Hormuz oil flows, forcing Thailand to borrow $12.2B, Indonesia to budget $5.9B in extra energy subsidies, and the Philippines to cut non-essential government spending. Malaysia's defence ministry was asked to contribute to savings. These fiscal pressures are directly competing with the AI and autonomous-systems procurement budgets that ASEAN militaries need to keep pace with Chinese and extra-regional capabilities.

ASEAN's February 2025 ADMM statement is voluntary, leaves 'human control' undefined, and has no verification or incident-reporting mechanism. Meanwhile Indonesia is fielding its first MALE drone, Malaysia has operationalised Anka-S, Thailand is building a Joint Capabilities Command for autonomous warfare, and the US-China autonomous-systems competition is physically operating in ASEAN waters. The window to establish binding norms before a miscalculation forces a fait accompli is narrowing rapidly.

The strength of US alliances that underpinned Southeast Asian security for decades is perceived to be eroding, with territorial disputes sharpening simultaneously. ASEAN states are accelerating domestic modernisation and diversifying partners (India BrahMos, Japanese OSA, Turkish Anka-S) precisely because the US security umbrella is seen as less predictable. This fragmentation risks interoperability gaps and competing AI/data architectures across partner forces.

The 2025 Cambodia–Thailand border clash demonstrated how rapidly kinetic escalation can occur between ASEAN members. Analysts warn the 'next round' with loitering munitions and ground robots could move from detection to detonation in seconds, leaving no space for human intervention — a danger compounded by AI-integrated LAWS becoming 'cheaper, more capable, and easier to field.'

China is systematically deploying AI across naval operations: the Qinzhou frigate carries an AI air-defence algorithm; submarine-launched AUVs perform seabed mapping; wave-powered autonomous USVs are proposed for persistent reef patrol; and AI-enabled 'intelligentized' seabed mines around the Paracel Islands are assessed as a near-plausible A2/AD evolution. China reportedly leads the US in AI drone-swarm applications. Twenty 'lily-pad' militarized installations on Woody Island and Antelope Reef confirmed in Jan 2026 satellite imagery.

China launched a commercial undersea AI data centre off Hainan Island, installed at 35m depth via submarine cables, adjacent to China's SSBN fleet at Longpo/Yulin bases. Analysts assess it provides 'operational and strategic advantages including proximity to major submarine cable routes.' Hainan already hosts naval bases, submarine facilities, and space-related operations — blurring civilian and military AI infrastructure.

A Reuters investigation (Oct 2025) documented coordinated influence activity — using local cutouts and cash incentives — targeting Filipino narratives on maritime claims, the US alliance, and public health. AI-scaled content generation and detection tools will intensify this information battleground. US officials named DeepSeek as supporting Chinese military and intelligence work while seeking chips through Southeast Asian intermediaries, raising compliance and counterintelligence risks for ASEAN markets.

Southeast Asian nations — Indonesia, Philippines, Vietnam — risk falling into a 'tiered security environment' where larger powers' AI-enhanced naval assets dominate surveillance, disrupt communications, or project force with fewer personnel and faster reaction times. Without interoperability frameworks or shared testing, technological inequality will exacerbate strategic asymmetry. ASEAN militaries 'will likely remain behind the stronger powers in autonomous systems for years to come.'

The February 2025 ADMM AI Joint Statement is voluntary and leaves 'operational definitions and verification open.' No binding human-control standard, no cross-border AI-incident hotline, and no moratorium on the riskiest LAWS deployments exist at the regional level. As Singapore's Foreign Minister noted at the UNSC: 'if strategic elements of deterrence are outsourced to AI systems, you are no longer dealing with another human being with inhibitions.' The gap between diplomatic statements and verifiable operational rules is widening.

China is developing wave-powered, near-silent AI-controlled USVs capable of year-long autonomous patrols around disputed reefs and islands at a fraction of conventional naval cost. Combined with the 'Underwater Great Wall' seabed sensor network and AI submarine-detection claims (Sep 2025 — 5% sub survival rate cited), China is systematically compressing the decision space available to ASEAN states and US partners. ASEAN has no doctrine or agreement to govern encounters with such systems.

The Cambodia–Thailand border clashes of 2025 demonstrated how quickly intra-ASEAN conflicts can escalate. Introducing loitering munitions and ground robots into similar future engagements could compress detection-to-detonation timelines to seconds with no space for political intervention. The February 2025 ADMM statement is voluntary and contains no operational definitions, verification, or incident hotline — leaving the region exposed.

Chinese AI models held ~30% of global usage share by end-2025, with deep penetration into ASEAN civil and emerging military AI infrastructure. Malaysia announced its sovereign AI stack uses DeepSeek (May 2025). US officials have simultaneously flagged DeepSeek as supporting Chinese military and intelligence work and seeking chips through Southeast Asian shell companies — creating acute counterintelligence and supply-chain compromise risks for ASEAN defence planners.

The strength of US alliances and security treaties that preserved regional stability for decades is assessed as fading precisely as territorial disputes sharpen. US INDOPACOM still describes AI as a 'force multiplier' but Trump-era deregulation and tariff shocks are undermining trust. The resulting strategic vacuum is accelerating uncoordinated national arms buildups, increasing the risk of miscalculation rather than deterrence.

A Reuters investigation (Oct 2025) documented coordinated influence operations targeting Filipino narratives on maritime claims, vaccines, and the US alliance using local cutouts and cash incentives. AI-scaled content generation will amplify such operations in the next cycle. ASEAN states lack coordinated AI-detection and counter-narrative infrastructure at comparable scale, creating exploitable cognitive vulnerabilities in frontline states most exposed to Chinese coercion.

China's South China Sea island outposts now provide 'all-weather, all-round situational awareness' per a July 2025 PLA article, integrating SIGINT, EW, and anti-ship/anti-air missile systems. The AI-enhanced Qinzhou-class frigate conducted combat drills in the SCS in 2025. China may have surpassed the US in AI-enabled drone swarm capability according to a Tamkang University analyst, and is developing wave-powered AI USVs for persistent SCS patrol — directly threatening ASEAN EEZ sovereignty.

Sustained Chinese Coast Guard and maritime militia operations at Second Thomas Shoal and Scarborough continue in 2026. A Reuters investigation (Oct 2025) documented coordinated AI-scaled influence operations targeting Filipino public opinion on maritime claims, vaccine narratives, and the US alliance using local cutouts. The US and Philippines expanded Balikatan 2026 with a heavy focus on contested-waters interoperability, raising Chinese counter-pressure.

Bloomberg ship-tracking data (May 2026) shows a sharp increase in Chinese vessel incursions within 10 nautical miles of Vietnamese-claimed waters compared to 2022–23. Vietnam's deterrence build-up — US patrol boats, coast guard cutters, diversified security partnerships — signals intent to contest Chinese encroachment, while analysts warn 'the odds of a confrontation are rising even as both countries seek to warm ties.'

Cambodia's use of fiber-optic drones against Thai forces in 2025 border clashes was Southeast Asia's first documented intra-regional UAS combat engagement. Thailand responded by establishing a UAS Warfare Center and a dedicated UAS battalion. The episode exposed the absence of any ASEAN-level norms or hotlines for drone incidents, and demonstrated how rapidly drone escalation can occur between member states.

US officials have named Chinese firm DeepSeek as supporting PLA military and intelligence work while seeking advanced chips through Southeast Asian shell companies. This raises acute compliance and counterintelligence risks for ASEAN economies — particularly Singapore, Malaysia, and Vietnam — that are actively courting cloud, semiconductor, and AI infrastructure investment, creating potential vectors for illicit technology transfer to Chinese defence programmes.

The February 2025 ADMM Joint Statement is voluntary and leaves operational definitions of human control and autonomous engagement undefined and unverifiable. Analysts at RSIS and the Lowy Institute warn this is 'a floor, not a ceiling.' Meanwhile, LAWS procurement is accelerating across the region through imported hardware pre-loaded with foreign doctrine — risking ASEAN states inheriting engagement thresholds and autonomy levels set by great-power suppliers, not by Southeast Asian strategic culture.

China's Wing Loong-2 UAVs conduct 32-hr continuous SCS patrols while deploying AI-guided UUVs (AJX002, UUV-300) and wave-powered USVs. Malaysia's ANKA-S, Indonesia's Elang Hitam, and US MQ-9s operated from the Philippines create overlapping autonomous ISR corridors with no agreed rules of engagement or hotline for AI incidents — raising misperception and accidental escalation risk.

The ADMM's February 2025 joint statement is voluntary and leaves 'human control' undefined. As SEA states acquire semi-autonomous systems, there are no enforceable minimum standards for weapons release authority, model-specific event logs, or cross-border AI incident hotlines — exactly the gap that Lowy Institute analysts warn could enable autonomous escalation with narrowing windows for political intervention.

The US expansion of MQ-9 Reapers across the Indo-Pacific creates a 'collective ISR web' that the PLA's SCSPI think tank characterises as a primary reconnaissance threat, reportedly countered with air defenses, jamming, and comparable Chinese unmanned systems. SEA states hosting US drone operations face dual pressure: enhanced domain awareness versus Chinese political and economic coercion.

Reuters documented coordinated AI-amplified influence operations targeting Filipino public narratives around South China Sea disputes in October 2025. With AI-generated disinformation scalable and cheap, all SEA claimant states face manipulation of domestic political support for maritime postures — a direct national security vulnerability feeding into crisis decision-making.

Only Vietnam has enacted binding AI legislation (effective March 2026). Indonesia is finalising mandatory requirements, Malaysia and Thailand have draft frameworks, while Cambodia, Laos, and Myanmar remain ungoverned. Divergent national rules create compliance arbitrage risks for dual-use AI systems and undermine the ASEAN-level soft-law architecture before it consolidates.

Thailand's restarted S26T submarine contract with China's CSOC (replacing German engines), China's dominance in commercial drone components, and the presence of Chinese-origin AI systems in SEA civil infrastructure create latent dependency vectors. As SEA militaries integrate AI into command systems, supply-chain provenance and backdoor risks in Chinese-origin hardware and software are insufficiently audited.

Reuters reported (Feb 2026) that a Chinese military long-range UAV (YILO4200) conducted 23+ flights over the SCS since Aug 2025, transmitting false transponder identities — including mimicking a sanctioned Belarusian cargo plane and a British Typhoon fighter jet. Operations were concentrated near the Paracel Islands and along Vietnam's coast. Analysts assess this as grey-zone electronic deception testing, directly relevant to future conflict escalation management and raising alarm in Hanoi and Manila.

China's 88.5-metre autonomous drone carrier — framed officially as a civilian ocean science vessel but capable of deploying 50+ aerial, surface, and underwater drones across a ~100-mile range — operated in a Philippine exclusive economic zone without authorisation in July 2025. Analysts warn that data collected could support PLA submarine navigation and anti-submarine warfare. Manila protested; incident intensified Philippine urgency to host US unmanned surface vessel infrastructure.

US officials identified Chinese AI firm DeepSeek as supporting PLA military and intelligence work while allegedly seeking advanced chips through Southeast Asian shell companies. This directly implicates ASEAN markets — particularly Malaysia, which in May 2025 announced its sovereign AI stack uses DeepSeek — in US export-control compliance and counterintelligence risk. The episode raises the stakes for all ASEAN states courting Chinese cloud and semiconductor investment alongside US alliance obligations.

The 2025 Cambodia–Thailand border skirmishes — resolved by direct leadership contact — illustrate how loitering munitions and ground robots could shrink human decision windows to seconds in future incidents. In the South China Sea, coast guard units, naval forces, maritime militia, and civilian traffic create an environment where AI-assisted autonomous systems risk triggering unintended escalation before political leaders can intervene. ASEAN has no binding instrument to govern such deployments as of June 2026.

A Reuters investigation (Oct 2025) detailed coordinated influence activity targeting Filipino public opinion on vaccines, South China Sea maritime claims, and the US alliance — using local cutouts and cash payments. Analysts assess that AI-scaled content generation and synthetic media will dramatically amplify such operations in the next cycle, eroding public support for Philippine security postures and US alliance legitimacy at a moment of peak SCS tension.

The US/Israel–Iran conflict disrupted Strait of Hormuz flows, forcing ASEAN energy-import-dependent economies into difficult trade-offs. Indonesia estimated USD 5.9B in additional energy subsidies; Thailand borrowed USD 12.2B; Malaysia's defence ministry was asked to contribute to savings; the Philippines told agencies to cut non-essential spending. The World Bank projects ASEAN growth slowing from 5.0% (2025) to 4.2% (2026), threatening the multi-year modernisation budgets that underpin AI-defence programmes across the region.