China Telecom's TeleAgent Tops 1.5M Users as AI Agents Go Mainstream

China Telecom's office AI agent, TeleAgent, has passed 1.5 million registered users and more than 200,000 daily active users, according to an announcement by company leadership at the opening of PT Expo China 2026. The milestone caps a remarkable run: the product counted just 27,000 registered users in mid-June, meaning adoption grew roughly fifty-five-fold in about four months.

Liu Guiqing, Executive Director, President and Chief Operating Officer of China Telecommunications Corporation, presented the figures at the expo's opening session on October 9. He described the milestone as evidence that autonomous AI agents capable of end-to-end task execution are moving from conceptual showcases into large-scale enterprise adoption. The product was first unveiled at the World Artificial Intelligence Conference (WAIC) 2026, reported Media OutReach Newswire via ASEAN Gazette.

The announcement positions TeleAgent as one of the fastest-growing workplace AI agents in the world, and one of the clearest real-world test cases for whether AI agents can hold millions of users in production environments rather than in demos.

From 27,000 to 1.5 million in four months

The velocity of TeleAgent's growth stands out even in a year crowded with agent launches. Few AI agent products disclose user numbers at all; fewer still show a curve that steep. According to the figures released at PT Expo China 2026, the registered-user base expanded from tens of thousands to well past a million in a single summer, with more than 200,000 people now using the agent every day.

The trajectory echoes a broader pattern across the agent ecosystem: enterprise AI products that can execute complete tasks — drafting documents, processing files, coordinating workflows — are pulling ahead of chat-style tools that require constant human prompting. The TeleAgent figures give that pattern a concrete anchor in the world's largest enterprise market.

A skill marketplace built almost entirely by users

One of the most striking details in the announcement concerns TeleAgent's skill marketplace, which has grown from roughly 40,000 to more than 50,000 skills. According to the company, 99 percent of those skills were built directly by end users solving their own productivity challenges, reported Newspatrolling.

That statistic matters. In agent platform economics, the moat is rarely the model — models get commoditized fast. It is the library of reusable, battle-tested workflows that users build and share. A 50,000-skill marketplace where the overwhelming majority of contributions come from the community suggests TeleAgent is becoming a genuine platform rather than just a product, with a flywheel where each new user potentially contributes skills that make the agent more useful for everyone else.

The skill-first design also reduces friction. Users drag and drop files into the TeleAgent workspace and let the agent handle end-to-end execution, with no complex server setup required for individual users, according to the release. The product runs natively on Windows, macOS, and leading Chinese enterprise operating systems including Kylin OS and UnionTech OS.

Sovereign deployment and the enterprise strategy

TeleAgent is developed by China Telecom AI Technology Co., known as TeleAI, and the company is pursuing a multi-edition strategy to cover the full enterprise market. An enterprise SaaS edition supports multi-device collaboration and scenario customization for teams. A private-deployment edition, released in mid-August, keeps models and data entirely within customers' local environments, reported Business Quest.

The private-deployment option is a significant design choice. Large enterprises and state-owned conglomerates — a core target market for TeleAgent — increasingly demand that AI systems run inside sovereign infrastructure. By shipping a local-deployment edition early, TeleAI is positioning itself to capture regulated and data-sensitive customers that cloud-only agent products cannot reach.

The product serves individual professionals, small and medium-sized enterprises, and large state-owned conglomerates, spanning the full range from solo productivity users to heavy enterprise deployments.

Why this milestone matters for the agent ecosystem

TeleAgent's 1.5 million users make it one of the largest disclosed user bases for a dedicated AI agent product anywhere. For an ecosystem that has been long on announcements and short on adoption data, hard numbers from a major carrier-scale rollout are genuinely useful evidence.

Three takeaways stand out. First, task execution beats conversation: the product's growth story is built around agents that do work end-to-end, not chatbots that wait for the next prompt. Second, user-built skills are the compounding asset: 99 percent user-generated contributions point to an ecosystem model that scales without proportional engineering investment. Third, deployment flexibility unlocks enterprise demand: the private-deployment edition shows that sovereignty and control remain prerequisites for the largest buyers.

Western agent companies have watched the US enterprise market closely for signs of what actually converts at scale. The push to measure whether agents behave reliably in production — the subject of recent agent-evaluation work — only matters once agents reach real scale, which is exactly what TeleAgent's numbers demonstrate. The figures, disclosed publicly at a major industry expo by the carrier's top operating executive, give the global agent industry a rare, concrete benchmark to measure against.

Whether the growth sustains into next year will depend on retention: daily actives above 200,000 suggest the product is finding real workflows, not just signups. But four months from 27,000 to 1.5 million is the kind of curve that makes the rest of the agent industry take notice.