Imagine a world where technology moves so fast, it’s like trying to catch smoke. That’s kind of what’s happening in Southeast Asia right now with artificial intelligence (AI). On May 24th, something big happened: ASEAN Secretary-General Kao Kim Hourn opened the new ASEAN-China Artificial Intelligence Industry Innovation Center in Beijing. Think of it like a grand opening for a super-smart technology hub, a “flagship” project under a major partnership between China and Southeast Asian nations from 2026 to 2030. This new center isn’t just about showing off fancy gadgets; it’s designed to bring people together to research, build, set standards, and help countries grow their AI capabilities. It’s a promise of a future where AI helps with everything from making things in factories to building smart cities and improving digital services. It’s a dazzling prospect, offering these nations a chance to leapfrog ahead technologically, much faster than they might have on their own.
But hold on a minute, because just like any fancy new tool, there’s a flip side. The timing of this center’s launch is pretty striking, because we’re entering a new era of AI, one that’s a bit scarier. We’re not just talking about AI that creates cool pictures or writes engaging text anymore. We’re talking about “agentic” AI systems. Picture these AIs as super-smart, independent problem-solvers. They can come up with their own plans, work with other AIs, and get things done with very little human help. It’s like having a team of tireless robots that can think and act on their own. Now, imagine these incredibly powerful, autonomous AIs in a region like Southeast Asia, which is a melting pot of different languages, where social media is everywhere, and where political feelings run high. This combination, unfortunately, makes the region a perfect breeding ground for a new, insidious kind of problem: agentic disinformation. This isn’t just your typical fake news; it’s fake news powered by super-smart AI that can adapt, personalize, and spread like wildfire, making it incredibly hard to distinguish from the truth.
We’ve already seen glimpses of this danger, and it’s a chilling preview of what could come if we’re not careful. For example, there’s been a surge of AI-generated videos and comments on Chinese social media, specifically targeting Singapore’s leaders with made-up stories about instability and economic collapse. It’s like a digital smear campaign, but instead of disgruntled individuals, it’s likely sophisticated AI churning out these narratives. (It’s crucial to understand that these recent disinformation attacks have absolutely no connection to the new ASEAN-China AI Center.) So, while the new Center promises groundbreaking advancements, it also highlights a classic dilemma in the Indo-Pacific: genuine opportunities for technological breakthroughs walk hand-in-hand with the risk that deeper integration of advanced AI tools, particularly those developed in China, could actually accelerate these dangerous agentic disinformation campaigns if the rules and safeguards don’t keep up with the technology’s rapid evolution. It’s a tightrope walk – the promise of progress on one side, and the peril of amplified manipulation on the other.
The Center’s work is built upon four main pillars, like the sturdy legs of a table ensuring its stability. First, it’s all about fostering shared research and getting AI working at an industrial scale, especially in manufacturing and those smart infrastructure projects we talked about. Second, they’re setting up channels for trade and investment, helping to grow the whole AI ecosystem. Third, there’s a big push for setting common standards and rules, so everyone is playing by the same book. And finally, the Center aims to build up the region’s capabilities through hands-on smart infrastructure projects across all the ASEAN member states, helping them to not just use, but truly understand and integrate AI. This isn’t completely new; it’s actually a stepping stone from an earlier initiative, the China-ASEAN Artificial Intelligence Application Cooperation Center, which was set up in Nanning back in September 2025. But this new Center shifts the focus, moving beyond just applications to practical, real-world industry rollout and deployment across borders, all aligned with the broader 2026-2030 Comprehensive Strategic Partnership Action Plan. So, in essence, the Center provides a clear path for ASEAN governments to speed up their digital transformation while also forging stronger, more interconnected relationships with Chinese AI developers, sharing knowledge and resources to build a more technologically advanced future together.
Yet, this very integration, this deep intertwining of technological development, also opens the door to significant governance risks – risks that our current policies and regulations simply aren’t equipped to handle. Imagine a situation where the rules of the road were written for horse-drawn carriages, and suddenly you have self-driving cars on the street. That’s the kind of gap we’re facing. Existing guidelines, like the ASEAN Guide on AI Governance and Ethics, and most national laws, are designed to manage AI systems where humans are still firmly in control. They don’t offer much guidance, however, on these autonomous “swarms” – networks of AI agents that operate independently and across national borders. The new Center, by encouraging joint research, sharing data, and building capacity, could inadvertently lower the barriers for the deployment of Chinese-developed agentic AI tools throughout the region if regional standards and safeguards don’t catch up at the same pace. This means that the very technical integration that promises so much economic prosperity and advancement could, ironically, reduce the barriers for influence operations to exploit these unresolved gaps in governance. Moreover, China’s approach to AI safety and governance, alongside state-led initiatives, heavily emphasizes “cyber sovereignty” and “civilizational AI values.” These values prioritize state control, data security, and maintaining social stability, and they are very likely to shape the joint standards and rule-making processes that emerge from this new Center. This means that as Southeast Asian nations embrace these integrated AI systems, they might also inadvertently adopt a governance framework that prioritizes state authority and control, potentially at the expense of individual freedoms or open information flows, leading to crucial strategic dilemmas.
The potential for these governance gaps to create serious strategic risks for the entire region is not just theoretical; it’s a very real and present danger. Imagine agentic disinformation – those highly sophisticated, AI-driven fake narratives – deepening political divides, eroding public trust in institutions like governments and media, and making it incredibly difficult to manage crises, whether it’s an incident in the South China Sea or a contentious national election. These aren’t just local problems; the weaknesses in ASEAN’s governance could have ripple effects far beyond Southeast Asia. It directly threatens the resilience of technology cooperation initiatives like those by the Quad (Australia, India, Japan, and the United States) and AUKUS (Australia, United Kingdom, and United States), making them more vulnerable. Furthermore, gaps in cybersecurity, a shortage of technical expertise, and insufficient monitoring create exploitable loopholes that malicious actors could easily leverage. Therefore, the new Center highlights a clear strategic tension: while deeper economic integration with China’s advanced AI promises great benefits, it also brings heightened risks of undue influence. To preserve “strategic autonomy” – the ability of Southeast Asian nations to make their own choices in the digital realm – proactive mitigation is absolutely essential. Without these safeguards, a phenomenon called “path dependence” will set in quickly. This means that the technical standards, data pipelines, and the very companies supplying these AI systems that are chosen now will become incredibly difficult and costly to change later. The early decisions made about governance will be absolutely crucial, determining whether the Center ultimately strengthens regional resilience or, sadly, locks in long-term vulnerability to external manipulation and control.
These risks aren’t some distant possibility; they are immediate and pressing, especially with tensions in the South China Sea showing no signs of easing and several ASEAN countries facing politically sensitive elections in the coming 18 months. Imagine agentic AI systems rapidly generating highly tailored disinformation in local languages, making it seem like there’s grassroots opposition where none truly exists, or manufacturing a false consensus around contested maritime claims. Once the technical standards, data pipelines, and the ecosystems of vendors are locked in through this new Center, reversing course would become significantly harder, if not impossible. Leaders are starting to acknowledge this. At the Shangri-La Dialogue in Singapore from May 29-31, these risks were directly addressed. Vietnam’s President To Lam, in his keynote address, emphasized that powerful technologies like AI and autonomous systems could amplify suspicion, enable information manipulation, speed up decision-making to dangerous levels, and increase the risk of miscalculations. Japan’s Defense Minister Koizumi Shinjiro echoed this, stating that social media and AI can be misused, shaking people’s decisions through disinformation. Singapore Defense Minister Chan Chun Sing highlighted the immense opportunities AI offers for national security, but stressed that no government wants an uncontrollable AI-enabled weapon system.
While these high-level speeches show a growing awareness, the discussions remained largely general, without concrete decisions or commitments on how to govern agentic AI. It’s like seeing the storm clouds gathering but not deciding to waterproof the house. ASEAN and its partners must move beyond mere awareness and translate these concerns into practical steps. Four key actions are crucial: Firstly, every new AI cooperation agreement needs to include “human-in-the-loop” and transparency rules, ensuring human oversight and accountability. Secondly, there needs to be joint monitoring of cross-border agentic activity, much like countries cooperate to monitor airspace. Thirdly, the ASEAN Guide on AI Governance and Ethics must be updated by the end of 2026 to specifically cover autonomous “swarms.” This isn’t just an update; it’s a complete rewrite for the modern age. Finally, Australia, Singapore, and Japan should launch smaller, focused initiatives to develop detection tools for agentic disinformation. These measures wouldn’t slow down innovation; in fact, they would ensure that the economic benefits of the ASEAN-China Center are accompanied by safeguards that protect regional stability and strategic autonomy. The ASEAN-China Artificial Intelligence Industry Innovation Center, in its essence, is neither purely good nor inherently bad. Its ultimate impact will largely depend on whether ASEAN and its partners treat the governance of this powerful technology with the same urgency and seriousness they apply to the economic opportunities it presents. The region must secure concrete commitments to prepare Southeast Asia for agentic AI, or risk allowing disinformation campaigns to exploit these emerging AI capabilities right alongside innovation. It’s a race against time, and the future of trust and stability in Southeast Asia hangs in the balance.