The Chinese Communist Party’s Engagement with Digital Human Technology

In recent years the rapid advancement of artificial intelligence has given rise to a category of technology that governments, media organizations, and commercial enterprises have rushed to adopt: the digital human. Known variously as virtual humans, virtual digital persons (虚拟数字人), or AI avatars, these computer-generated figures are designed to interact with real people through synthesized speech, facial animation, and increasingly naturalistic gesture. Among the world’s governments, the Chinese Communist Party has pursued this technology with exceptional breadth and institutional commitment, deploying digital humans across party education, state media, government services, cultural heritage, and ideological dissemination. What began as isolated experiments around 2018 has matured into a nationwide ecosystem underpinned by dedicated Five-Year Plan provisions, municipal funding programs totaling tens of billions of yuan, sector-specific national standards, and, as of 2026, the world’s first draft regulation governing digital human information services.

The party’s earliest and most symbolically charged application of digital human technology has been the recreation of revolutionary martyrs and historical figures for patriotic education. The most prominent example emerged during the CCP’s centenary celebrations in July 2021, when Bilibili creator Hu Wengu (大谷Spitzer) collaborated with People’s Daily to produce a viral video restoring and animating photographs of revolutionary figures including Li Dazhao, Xia Minghan, Fang Zhimin, and Zhao Yiman, using Alibaba’s GPEN face restoration algorithm and Megvii’s RIFE video interpolation framework. The video trended to the top position on Weibo and was endorsed by People’s Daily with the caption: “The China of today is as you wished.” The company Digital Xusheng (数字栩生), working with Beijing Institute of Technology and the Central Academy of Drama, subsequently produced high-fidelity 3D digital human reconstructions of Li Dazhao, Lu Xun, and Mei Lanfang for cultural education deployments. By 2025, Chengdu Neusoft University’s AI Center had created a digital human recreation of the revolutionary martyr Gan Xiyu, deployed at Dujiangyan Martyrs’ Cemetery in Sichuan, where visitors can engage in two-way spoken dialogue with the virtual figure about his life and sacrifice. The CCP First Congress Memorial Hall in Shanghai has taken an even more immersive approach, using multi-screen holographic theater to recreate the seven-day founding congress of 1921, while its “Digital First Congress” (数字一大) app employs digital twin technology to let users explore the historical setting.

The integration of digital humans into museums and exhibition spaces has extended well beyond isolated pilots. The flagship deployment is Shen Xiaoyi (申小伊), designated the “No. 001 AI digital human guide” at the CCP First Congress Memorial Hall, launched on June 3, 2021 and developed by SenseTime (商汤科技). Trained on more than 10,000 party history questions, Shen Xiaoyi has served tens of thousands of visitors and is also accessible through the Digital First Congress app, which was co-produced with China Unicom and debuted at the World Artificial Intelligence Conference in 2023. The platform was named among Shanghai’s first batch of metaverse application scenarios and has been deployed remotely to sites in Kashgar in Xinjiang and Nanning in Guangxi. At the National Museum of China in Beijing, the digital guide Ai Wenwen (艾雯雯) debuted on July 22, 2022, developed by Fantawild Digital (凡拓数创) with support from Beijing University of Posts and Telecommunications. The same company created Jing Hui (京慧) for the Capital Museum. Additional verified deployments include a digital guide at the Puppet Emperor’s Palace Museum in Changchun, Baidu’s Wen Yaoyao (文天天) virtual cultural relics interpreter, and the Aikesheng (爱可声) AI digital guide system deployed across more than twenty museums since July 2024, with plans to expand to one hundred.

In the realm of state media, the use of AI-generated anchors has evolved from a novelty into standard practice across all tiers of the Chinese broadcast system. Xinhua News Agency debuted what it described as the world’s first AI news anchors on November 7, 2018 at the 5th World Internet Conference in Wuzhen, developed in partnership with Sogou. The Chinese-speaking version, Xin Xiaohao (新小浩), was modeled on the human anchor Qiu Hao, and an English-speaking counterpart launched simultaneously. By May 2020, Xin Xiaohao had produced more than 13,000 news reports totaling over 35,000 minutes of broadcast time. Subsequent iterations included the first female AI anchor, Xin Xiaomeng (新小萆), in February 2019, a Russian-speaking anchor launched at the St. Petersburg International Economic Forum in June 2019, and the first 3D AI anchor, Xin Xiaowei (新小微), unveiled on May 21, 2020. In June 2021, Xinhua and Tencent’s NExT Studios created the digital reporter Xiao Zheng (小诬) to cover the Shenzhou-12 crewed space mission. People’s Daily launched its first AI anchor, Guo Guo (果果), on May 25, 2019, followed by the interactive anchor Ren Xiaorong (任小融) in March 2023 during the Two Sessions. CCTV has deployed numerous virtual presenters, including the Lingyu (聆语) AI sign language anchor built by Tencent Cloud for the 2022 Beijing Winter Olympics, with a vocabulary exceeding 1.6 million words, and Xiao C (小C), the first digital person to conduct live interviews with National People’s Congress representatives. By 2025, AI anchors had spread to provincial and county-level media centers, with Hangzhou achieving the first fully AI-anchored newscast during Spring Festival 2025, integrating the DeepSeek-V3 large language model. The technology partner ecosystem is well-defined: Sogou (now owned by Tencent) for Xinhua, iFlytek (科大讯飞) for CCTV voice synthesis, Baidu for CCTV.com digital humans, and Tencent for sign language and reporting applications.

The party’s promotion of digital human technology extends well beyond its own immediate operational needs. The Chinese government has constructed a layered policy architecture that channels public funding, sets industrial targets, and establishes technical standards for the sector. At the national level, the 14th Five-Year Plan Outline adopted in March 2021 designated virtual reality and augmented reality as one of seven core digital economy industries. The 14th Five-Year Plan for Digital Economy Development, issued as State Council Document Guofa 2021 No. 29 in December 2021, called for strengthening research in human-like intelligence, natural interaction, and virtual reality. The MIIT Action Plan for Virtual Reality and Industrial Applications covering 2022 to 2026, jointly issued by five ministries in November 2022, set a target of 350 billion yuan in VR industry output by 2026. Most directly relevant, the MIIT Metaverse Industry Innovation Development Three-Year Action Plan, issued as Document Gongxinting Lianke 2023 No. 49 in August 2023, explicitly names digital human guides, shopping assistants, and customer service agents as priority applications. The 15th Five-Year Plan Outline adopted in March 2026 further elevates AI to a higher strategic priority, calling for comprehensive implementation of an “AI+” action and breakthroughs in multimodal AI, intelligent agents, and embodied intelligence, all technologies that directly underpin digital human systems.

Municipal governments have committed the most specific funding. Beijing issued China’s first dedicated digital human industry policy in August 2022, targeting an industry scale exceeding 50 billion yuan (approximately seven billion US dollars) by 2025, with goals of cultivating one to two enterprises exceeding five billion yuan in revenue and ten exceeding one billion. Shanghai published its Metaverse New Track Action Plan in July 2022, targeting 350 billion yuan in related industry scale and featuring a dedicated Digital Human Comprehensive Enhancement Project alongside a state-backed metaverse industry fund. Shenzhen’s “20+8” Industrial Cluster Strategy, launched in June 2022, includes virtual digital humans as a strategic emerging industry, offering up to five million yuan per research and development project. Guangzhou’s Nansha district offers up to 200 million yuan per platform for major metaverse research facilities.

The standards infrastructure has developed in parallel. The MIIT industry standard YD/T 4393, published in July 2023, established the first reference framework for virtual digital human metrics, drafted by CAICT with participation from Tencent, Baidu, and SenseTime. The broadcasting standard GY/T 411-2024, issued in November 2024, set technical requirements for digital humans in media. China’s first national standard specifically for digital humans, GB/T 46483-2025, was released in November 2025 covering customer-service virtual digital humans. Internationally, China led the development of ITU-T Recommendation F.748.15, titled “Framework and metrics for digital human application system,” which was frozen in January 2022, making it one of the earliest international standards in the field. At the First China Digital Human Conference in September 2024, MIIT disclosed that China had 1.144 million digital-human-related enterprises, with more than 174,000 new registrations in the first five months of 2024 alone.

In the context of governance and public services, digital humans are being deployed as virtual civil servants to handle routine inquiries and administrative tasks. Shanghai’s Pudong district launched its intelligent government service digital human Xiao Pu (小浦) in June 2024, handling company registration changes, public health permits, and transport permits, and completing more than 1,300 cases during its trial operation. Nanjing’s government service digital human Ning Xiao Zhu (宁小助) processes 32,000 daily consultations and diverts 65 percent of human agent workload, earning recognition as one of the Top 10 National Intelligent Government Service Cases by the State Council’s E-Government Office. Wuxi launched its first batch of more than twenty government service digital humans in April 2025 and subsequently signed a strategic partnership with Xiaoice Company. Yicheng county in Hubei became one of China’s first AI pilot counties in July 2024, deploying Huawei Cloud technology to cover 304 high-frequency government service items. Additional deployments have been documented in Beijing’s Fengtai and Changping districts, Inner Mongolia’s Tongliao, and Guangzhou’s Tianhe district. The regulatory infrastructure for these deployments is also emerging: Shenzhen’s Futian District issued China’s first interim measures for managing “government auxiliary intelligent robots” in September 2024, requiring each AI digital employee to have a designated human supervisor.

In education, digital human teachers are being developed to supplement rather than replace traditional instruction. Beijing No. 4 Middle School’s Xiongan Campus deployed an AI digital teacher called Xiao You, developed by 4UTech, for digital practice classes. Zhejiang University pioneered an AI-enhanced ideological education reform in the 2024–2025 academic year, creating digital human lecturers modeled after real professors with support from the Chaoxing (超星) platform, serving more than 5,000 freshmen. Tianjin University deployed a digital twin of Professor Liu Yanli in a power systems course beginning April 2025, developed with iFlytek, capable of generating lecture videos and personalized assignments. East China Normal University developed digital human “students” to train pre-service teachers, a project selected by the Ministry of Education as an “AI + Higher Education” typical case. New Oriental partnered with SenseTime’s Ruying (如影) platform to generate batch AI teaching videos. The Ministry of Education released its Guidelines for Generative AI Use in Primary and Secondary Schools, 2025 Edition, in May 2025, explicitly mentioning digital humans as an AI tool for teaching, while iFlytek has partnered with 27 universities for the Ministry’s first Generative AI Education Model projects.

Cultural heritage preservation and tourism promotion represent one of the most actively deployed application areas. Tencent has emerged as the dominant technology partner, with its collaboration with the Dunhuang Research Academy producing the Digital Sutra Cave (数字藏经洞), described as the world’s first game-technology-powered super-temporal participatory museum. Virtual cultural ambassadors Jiayao (伽瑶) and Tianyu (天妤) have been created for the project, with the latter’s short drama series exceeding 300 million video views. The broader Digital Dunhuang resource library has been accessed by 23 million users from more than 78 countries. Zhonghua Book Company created a 3D ultra-realistic digital Su Dongpo in August 2022, which appeared on CCTV’s Chinese Poetry Conference in 2023 for real-time holographic interaction with the live audience. At Tengwang Pavilion in Nanchang, a virtual guide named after Tang Dynasty poet Wang Bo narrates history and customizes tour routes. The Forbidden City has built a digital cultural relics library of more than 150,000 high-resolution items and deployed the AI tour guide Mr. Fu (福大人). The company Aikesheng Digital Cultural had deployed AI digital human guide systems in more than one hundred museums and scenic sites by March 2026, supporting over ten major languages. Provincial tourism AI systems include Guizhou’s Huang Xiao Xi (黄小西) and Shanghai’s Hu Xiao You (沪小游) smart tourism platform.

The development and deployment of digital humans has been accompanied by a rapidly evolving regulatory architecture. The Provisions on the Administration of Deep Synthesis Internet Information Services (互联网信息服务深度合成管理规定), jointly issued by the Cyberspace Administration of China, MIIT, and the Ministry of Public Security on November 25, 2022 and effective January 10, 2023, directly addresses digital human technologies by regulating face generation, voice synthesis, and virtual scene creation, mandating content labeling, real-name authentication, and algorithm filing. The Interim Measures for the Management of Generative Artificial Intelligence Services (生成式人工智能服务管理暂行办法), issued July 10, 2023 by seven agencies and effective August 15, 2023, regulate all generative AI content including digital human outputs. Subsequent milestones include the AI Safety Governance Framework Version 1.0 in September 2024 and Version 2.0 in September 2025, both issued by TC260; mandatory Measures for Labeling AI-Generated Content with national standard GB 45438-2025 effective September 2025; and the Cybersecurity Law Amendments effective January 2026, which for the first time elevated AI governance provisions into national law.

Most significantly for the digital human sector specifically, two draft regulations emerged in rapid succession. The Provisional Measures on Human-like Interactive AI Services, published for comment in December 2025, govern AI systems that simulate human personality and emotional interaction. The Draft Measures for the Administration of Digital Virtual Human Information Services, published on April 3, 2026 with a comment period closing May 6, 2026, constitute China’s first regulation specifically targeting digital humans, mandating continuous on-screen labeling indicating digital human status, requiring consent for the use of real persons’ likenesses, and establishing protections for minors against emotional dependency. China’s first national standard for virtual digital humans, GB/T 46483-2025, was released in November 2025. A comprehensive AI Law remains in the research and drafting phase, with implementation widely expected to follow after 2027.

The integration of digital humans into party affairs extends to the dissemination of current party ideology. The Xuexi Qiangguo (学习强国) app, launched January 1, 2019 and developed by Alibaba Group, had surpassed 100 million users by April 2019, delivering Xi Jinping speeches, party history courses, and political quizzes through a gamified study-points system. In May 2024, the China Institute of Cybersecurity Affairs, operating under the Cyberspace Administration of China, announced the Xue Xi (学习) AI chatbot, developed by Tsinghua University and trained on seven databases including one dedicated to Xi Jinping Thought. Baidu launched a Xuexi Qiangguo Official Document Assistant on its Ernie Bot platform in late 2024, enabling civil servants to search Xi Jinping’s articles and quotes to generate government documents. iFlytek partnered with Xuexi Qiangguo in September 2024 to release AI writing tools and smart earbuds for meeting transcription. People’s Daily launched its Easy Write (写易) AI tool in early 2024, described as a mainstream values large language model using only party-vetted sources. China Youth Daily has begun deploying AIGC and digital human technologies to explain Xi Jinping Cultural Thought using youth-oriented language and formats, representing the closest integration of digital human presentation technology with ideological dissemination, though the scale of these deployments remains limited compared to the text-based AI platforms that carry the primary load of party study content.

The international dimensions of China’s digital human development are real but narrower than the domestic ecosystem might suggest. The strongest evidence of international projection comes from state media partnerships. In April 2019, Sogou signed a formal agreement with Abu Dhabi Media in the UAE to develop the world’s first Arabic-speaking AI news anchor, in a ceremony witnessed by the UAE’s Minister of Artificial Intelligence. In June 2019, Xinhua, the Russian news agency ITAR-TASS, and Sogou launched a Russian-speaking AI anchor at the St. Petersburg International Economic Forum, timed for the 70th anniversary of China-Russia diplomatic relations. CGTN developed an interactive Ask AI Confucius project answering questions in eight languages, deployed at the 2025 Confucius Cultural Festival and winning an Asia-Pacific Broadcasting Union Prize for Digital Innovation. At a November 2025 global Chinese language summit, China unveiled the AI virtual teacher Xiao Yu for potential use across Confucius Institutes. The most commercially significant international channel, however, is cross-border e-commerce: Xiaoice partnered with Quantum Planet to deploy AI livestream hosts supporting 129 languages for Chinese cross-border merchants, while companies like BocaLive offer AI avatars in 29 languages for merchants in Yiwu and elsewhere. Companies including SenseTime, with offices in Singapore, Japan, Saudi Arabia, and the UAE, and iFlytek, which has invested 400 million Hong Kong dollars in a global headquarters in Hong Kong, maintain international presence, but specific digital human technology exports beyond the media partnerships and e-commerce tools remain limited.

The Chinese Communist Party’s engagement with digital human technology thus represents a comprehensive and systematically supported effort spanning party education, state media, governance, cultural promotion, and ideological dissemination. The effort is documented through official policy document numbers, named technology deployments with identified corporate partners, and specific funding targets at both national and municipal levels. The regulatory trajectory is particularly notable: China has moved from broad AI ethics principles articulated in 2019 to the world’s first dedicated digital human draft regulation in April 2026, a seven-year evolution from soft guidelines to binding sector-specific rules. The scale of state media adoption alone, with Xinhua’s AI anchors producing over 13,000 reports and the technology spreading to county-level media centers, demonstrates that what began as experimentation has become institutionalized practice. The policy infrastructure encompasses at least four national-level action plans, dedicated municipal programs in five major cities, three tiers of technical standards from industry to national to international, and the world’s earliest ITU recommendation on digital human systems. As the technology continues to evolve under the impetus of generative AI and large language models, and as the regulatory framework moves from draft to enforcement, the party’s use of digital humans will continue to expand and reshape the landscape of political communication and public engagement in China.