Ant Group Begins Internal Testing of 'Lingguang' AI Agent

Ant Group has started small-scale internal testing for a new version of its AI product 'Lingguang', which is benchmarked against Meta's Muse personal agent. This update shifts Lingguang from a basic chat and application-generation assistant into an autonomous AI-native personal agent. This development highlights the broader industry push toward autonomous personal agents capable of handling complex, long-term tasks rather than just simple conversational queries. By integrating long-term memory and cross-platform execution, it points toward more practical, everyday AI integration. The upgraded Lingguang features long-term memory and the ability to autonomously advance tasks by invoking Skills and the Model Context Protocol (MCP). Initial testing use cases focus on complex scenarios such as tracking market fluctuations, managing DingTalk check-ins, and rescheduling travel routes during disruptions.