Ant Group unveils 'Lingying' security brand and ASL protocol for agent safety

At the 2026 Outer Bund Summit on September 11, Ant Group's chief security officer for big security, Chen Liang, said that trustworthiness is a prerequisite for the large-scale application of AI. He noted that capability determines what AI can do, while trust determines what society is willing to delegate to AI. As agents begin to call tools, operate devices, and participate in transactions on behalf of users, the industry must address not only whether AI is intelligent, but also intent understanding, authorization boundaries, controllability, and traceability.

Chen said Ant is extending its security capabilities from the mobile internet era to the agent internet era, building an end-to-end security chain covering terminal entry points, agent identity, permissions, behavior, transaction risk control, and post-hoc auditing.

Wan Xiaofei, head of terminal security at Ant Group, disclosed that the GPASS brand has been upgraded to "Lingying," and GOS has been upgraded to "Lingying AOS," shifting from individual terminal security capabilities to underlying infrastructure for intelligent terminals. Lingying integrates trusted execution, data protection, and runtime environment security into the terminal and operating system layer, collaborating with device makers and supply chain partners to embed security into devices such as phones, AI glasses, and earbuds.

Ant also launched the ASL (Agent Security Trusted Interconnection Protocol), which focuses on security collaboration between agents, addressing identity trust, intent trust, and data trust. The protocol has been deployed in cross-platform agent scenarios such as 1688's e-commerce procurement and is used internally in Alipay. Chen emphasized that building security infrastructure for the AI-driven economy requires joint efforts from standards bodies, industry partners, platforms, and developers.