360 Launches NanoWork Enterprise Agent Platform, Zhou Hongyi Says AI Is Moving From 'Answering' to 'Working'

On July 28, 360 Group founder Zhou Hongyi officially launched NanoWork (纳米Work), a next-generation enterprise agent work platform positioned as infrastructure for business AI transformation. The platform targets business owners, entrepreneurs, and frontline teams, aiming to bridge the gap between AI's ability to answer questions and its ability to execute real work.

Zhou emphasized that NanoWork is not a simple wrapper around any single agent product, but rather an effort to hide complex technical capabilities and transform advanced AI into practical tools accessible to ordinary employees. His keynote framed the shift as AI evolving from "being able to answer" to "being able to work."

Before the launch, 360 treated itself as a testbed. The company deployed various agents including OpenClaw across 1,000 internal business scenarios, collected over 56,000 user feedback entries, and iterated through 166 versions over five months before finalizing NanoWork. This extensive internal testing reflects 360's approach to validating enterprise AI products at scale before public release.

Concurrent with the launch, 360 initiated the NanoWork Box national channel partner recruitment program, signaling its strategy to build an enterprise AI service ecosystem through partner networks. The channel approach aims to reach small and medium enterprises that may lack in-house AI expertise.

Zhou had previously previewed NanoWork at the ISC conference, where he argued that large language models are evolving from conversational AI toward what he called the "era of tens of billions of agents." NanoWork represents the commercial productization of that thesis, packaging agent capabilities into a deployable enterprise work platform.

The journey from internal testbed to commercial launch mirrors a growing pattern in China's enterprise AI market: organizations first validate products through extensive internal deployment, then leverage channel networks for market expansion. This model helps reduce trial-and-error costs for enterprise buyers but makes product maturity heavily dependent on the breadth of internal testing scenarios.