Vietnam's ruling Communist Party is drafting a plan to build a network of at least 1,000 influencers and 5,000 AI experts by 2030, use Vietnamese-built AI tools to lead social discussion, and ensure that 80 percent of online content in Vietnamese is positive while 90 percent of guideline-violating content is removed within 24 hours, according to an unpublished April document reviewed by Reuters.
The plan does not replace censorship. It sits on top of it. Vietnam already blocks foreign news outlets, prosecutes online critics, and requires 24-hour takedowns of content deemed false or anti-state. What the party's propaganda committee is now proposing is an additive layer: state-directed persuasion that looks culturally authentic because it uses formats and figures the audience already trusts. The draft describes the goal as building ideological immunity for society, which is a revealing phrase. It frames the entire Vietnamese online information environment as a body that needs to be inoculated against outside ideas, with AI as the delivery mechanism and influencers as the immune cells.
The operational mechanics are direct and practical. Recruited influencers will post approved material, promote official activities, and receive perks such as sponsored trips in exchange for participation without formal financial compensation. The Communist Youth Union recently invited Vietnamese influencers on a study trip to China, continuing a pattern of visits designed to familiarise content creators with the China model of state-aligned media culture. That reference matters. China's approach to digital propaganda has evolved from blunt censorship toward a sophisticated ecosystem of state-adjacent content creators, patriotic livestreamers, and algorithmically amplified positive coverage. Vietnam is looking at that ecosystem and deciding to replicate core elements of it with Vietnamese characteristics.
The comparison to China is instructive but not identical. China operates at scale that Vietnam cannot match, with a domestic social media ecosystem in WeChat, Weibo, and Douyin that the state can regulate directly. Vietnam's population is deeply embedded in Facebook, TikTok, and YouTube, all foreign platforms with their own moderation policies and political pressures. That creates a harder problem. The party wants to flood those platforms with positive content while also requiring them to take down negative content within 24 hours. Platforms have resisted similar demands elsewhere, and the requirement to be both amplifier and censor simultaneously puts them in an uncomfortable structural position. Vietnam has previously blocked Facebook temporarily over compliance disputes, suggesting the relationship is coercive rather than collaborative.
For SF readers, the policy document matters beyond its political context. The same AI media tools that startups use for localization, content at scale, social media scheduling, and influencer management are now being explicitly adopted as state-influence infrastructure. Vietnam's plan involves AI tools built by Vietnamese tech companies to lead social discussion, which means the government wants domestic AI startups to build the instruments of ideological management. That creates a business opportunity inside a difficult ethical environment. A startup that wins a contract to build AI content generation tools for the Vietnamese government does not just face regulatory risk. It faces the question of whether its product is being used to suppress democratic speech at scale.
That question is not hypothetical for founders selling into Southeast Asian markets. Vietnam is not alone. The Reuters document describes a strategy that multiple one-party systems in the region have studied or partially implemented. The tools involved, short-form video generation, social media monitoring, influencer management platforms, and content moderation AI, are exactly the products many AI startups are building for commercial advertising and brand management clients. The political and commercial use cases share infrastructure, which means every startup selling into government or large enterprise in politically sensitive markets needs a clear-eyed policy on how its tools can be used.
The broader trend here is the creator-economy playbook moving into state hands. Governments have watched influencers build trust, drive behavior, and shape narratives with more cultural agility than traditional broadcast media ever could. The natural conclusion for authoritarian systems is to recruit those same tools and those same figures, and to use AI to scale the output while managing the message. Vietnam's 2030 targets are ambitious given the pace of platform change and the fundamental challenge that state-aligned content often loses authenticity when audiences realize it is coordinated. But the direction of travel is clear. AI is making it cheaper, faster, and more culturally targeted for any organization, commercial or governmental, to flood an information environment. Startups building those tools need to decide early how that capability will or will not be used.
Also read: AWS overheating in Virginia is a reminder that cloud reliability is now a physical problem • SoftBank's OpenAI margin loan is the most leveraged bet in the history of artificial intelligence • Nvidia's $2.1 billion bet on IREN shows miners with power are becoming AI infrastructure
