Skip to content
aiqbench takeAI Policy

aiqbench analysis

This regulatory update from China's MIIT mandates rigorous risk assessments and unique identity codes for AI agents deployed on consumer terminal devices like smartphones and wearables. The policy effectively treats agent terminals as critical infrastructure requiring specific adaptation verification and proactive anti-fraud evaluations before network access is granted. This marks a significant shift in how governments are approaching the governance of autonomous software agents within existing telecommunications ecosystems.
Sources covering36Kr, WISeKey, Cisco Blogs, Quantum Zeitgeist

MIIT: Strengthening Network Access Management for Agent Terminal Devices

36Kr1 min read

36Kr reports that the MIIT has issued a notice on strengthening network access management for agent terminal devices. The notice calls for improvements to the network access management of agents, including conducting risk assessments and refining the telecommunications equipment network access testing mechanism with detailed specific requirements. Mobile phones, tablets, wearable devices, and other agent terminals pre-installed with agents must undergo agent adaptation verification in accordance with relevant standards during network access testing. Risk safety assessments for fraud involving agent terminals and proactive evaluations of anti-fraud capabilities will be conducted in compliance with the law, implementing important communication work requirements. Agents used on terminals should be equipped with a unique identity code referencing the national standard series on agent interconnection. Terminal production enterprises must ensure consistency between agent terminals before and after obtaining network access licenses. If significant changes occur to agent terminals, terminal production enterprises must report them to the MIIT and conduct network access testing as required.

Read the full article at 36Kr

Related coverage