China’s rapidly growing artificial intelligence (AI) ecosystem gained a powerful new player this week. Beijing-based startup Manus officially entered the spotlight after securing major backing from government entities and industry giants. The company claims it has built the world’s first general-purpose AI agent capable of autonomous decision-making. Chinese regulators have already approved its core product, Monica, as a registered generative AI application.
With its technology, user base, and political support all scaling simultaneously, Manus now positions itself at the forefront of China’s AI race — and possibly the global AI landscape.
Monica: The World’s First General AI Agent?
Manus unveiled its flagship product, Monica, earlier this year. The company designed Monica to go beyond typical AI assistants. Unlike ChatGPT or Siri, Monica doesn’t just respond to commands — it initiates tasks, learns user behavior, and makes proactive decisions across digital platforms.
Manus engineered Monica to complete a wide variety of tasks without direct instruction. For example, Monica can manage your calendar, execute trades, answer emails, and even set up business meetings on your behalf — all based on context it learns autonomously.
According to Manus founder and CEO Zhang Lin, Monica integrates deeply with various online systems. “We trained Monica on hundreds of millions of real-world interaction scenarios,” Zhang said during a recent press conference. “She can negotiate, manage logistics, and operate as a full-service virtual agent.”
This level of autonomy sets Monica apart from current large language models (LLMs), which require explicit prompts and human guidance. Zhang believes this advancement pushes AI closer to Artificial General Intelligence (AGI), where machines can learn and apply knowledge across any task.
China’s Government Fast-Tracks Monica’s Approval
China’s Ministry of Industry and Information Technology (MIIT) approved Manus’s AI model under its latest generative AI regulatory framework. This approval allows Manus to legally distribute Monica in China — a milestone few startups have achieved. So far, only a handful of domestic AI platforms have passed the rigorous compliance process.
The government prioritized Monica’s application, citing its national security relevance and potential economic impact. Manus also joined an elite group of government-partnered AI firms that operate under the “trusted innovation” umbrella — a designation that grants it access to public contracts, subsidies, and enterprise-scale data.
Industry analysts noted that the government’s decision signals a deeper commitment to homegrown AI platforms, especially amid increasing tensions with Western tech powers. Manus, backed by several prominent venture firms and corporate alliances, now serves as a flagship model for China’s AGI aspirations.
Strategic Alliances with Alibaba and Qwen AI
Manus didn’t build its platform in isolation. It partnered with Alibaba, integrating its AI model with Alibaba’s open-source Qwen models. Qwen, developed by Alibaba’s DAMO Academy, serves as a foundation for many Chinese LLMs. By aligning with Qwen, Manus gained access to cutting-edge model architecture, robust training infrastructure, and a shared ecosystem.
Alibaba also helped Manus secure computing resources through its Tongyi Qianwen cloud platform. This alliance enabled Manus to process billions of tokens and train Monica across highly diverse datasets. As a result, Monica developed advanced multilingual and contextual reasoning skills far beyond typical chatbot systems.
Zhang confirmed the collaboration during a state AI conference. “We didn’t reinvent the wheel. We collaborated with national champions to focus on what we do best — building AI agents that act,” he said.
2 Million Users — and Growing
Despite launching in closed beta, Manus has already accumulated a waiting list of over 2 million users. Corporate clients, entrepreneurs, and developers have flooded Manus with early access requests. Users hope to integrate Monica into business workflows, customer service pipelines, and logistics operations.
Early reports suggest that Monica performs particularly well in financial modeling, supply chain coordination, and automated customer support. Several startups in fintech and e-commerce have already begun internal trials.
Zhang stated that Manus will roll out Monica in waves, starting with enterprise customers and government institutions. By mid-2025, the company aims to onboard consumer users through a mobile-first platform that supports smart homes, digital payments, and lifestyle management.
A Strategic Response to U.S. AI Dominance
China’s accelerated support for Manus does not exist in a vacuum. The move reflects a strategic response to perceived U.S. dominance in generative AI, led by companies like OpenAI, Anthropic, and Google DeepMind.
In 2024, U.S. export controls restricted the sale of high-end AI chips to Chinese companies. That policy shift forced China to double down on domestic AI innovation, both in hardware and software. In this context, Manus symbolizes a key step in reclaiming AI leadership through indigenous solutions.
Professor Liu Xiaoyu, a technology policy expert at Tsinghua University, described the Manus rollout as “a national moment.”
“Manus brings China closer to AGI. The fact that it does so with regulatory approval, industrial integration, and national support makes it a strategic pillar,” she said in an interview with China Daily.
Challenges Ahead
Despite the excitement, Manus still faces critical challenges:
- Compute Constraints: Even with Alibaba’s support, China continues to face chip shortages. Manus needs vast GPU clusters to maintain Monica’s performance as user load increases.
- Privacy and Ethics: Monica’s autonomy raises new questions about digital responsibility. Users entrust it with financial, legal, and personal decisions — which could lead to disputes or misuse.
- Global Expansion Limits: International launch remains unlikely due to data compliance laws. Western markets may restrict Monica unless Manus complies with localized AI governance frameworks.
- User Trust and Control: While Monica acts autonomously, some users express concern about losing control. Trusting an AI to make decisions without real-time approval introduces psychological and operational risks.
Manus acknowledges these hurdles. The company recently hired former Huawei security executive Tang Yiming to lead its AI ethics board. Tang will oversee transparency, user consent, and data protection initiatives as Monica scales.
Future Roadmap
Looking ahead, Manus plans to introduce Monica Pro, a premium enterprise version with integration support for CRMs, ERPs, and financial software. The company also intends to release developer APIs by Q3 2025, allowing external engineers to build on top of Monica’s architecture.
To handle the growing workload, Manus will expand its team from 150 to over 400 employees by year-end. Most hires will focus on natural language processing, reinforcement learning, and data security.
CEO Zhang Lin remains optimistic about the future. “We don’t just want Monica to be smart. We want her to be reliable, secure, and capable of working alongside every human — whether it’s in a hospital, an office, or a smart home,” he said.
Conclusion
Manus has emerged as one of China’s most ambitious and well-positioned AI startups. With its Monica AI agent, government endorsement, and powerful partnerships, the company stands at the center of China’s strategy to lead the world in AI development.
While challenges remain — especially in scalability, ethics, and international acceptance — Manus continues to push the boundaries of what autonomous AI can achieve. Whether Monica truly represents the dawn of AGI or not, she has already redefined what’s possible for AI agents in real-world environments.
As China accelerates its digital ambitions, Manus shows how quickly a startup can go from stealth mode to national priority — and possibly global disruption.