The China Academy of Information and Communications Technology (CAICT) has formally initiated the development of a comprehensive set of standards for Intelligent Assistant Agents, known as Claw, with the objective of fostering the establishment of a robust and relevant standard framework. At present, the Trustworthy Capability Requirements for Intelligent Assistant Agents (Claw) Products are open for contributions from industry organizations and experts. These contributions will help define the criteria necessary to ensure the quality controllability and behavioral trustworthiness of Claw products. Concurrently, the registration process for the inaugural batch of trial evaluations has also commenced.

The trial evaluation, aimed at assessing the trustworthy capabilities of Claw products, is slated for the end of March 2026. Following this evaluation, the first batch of results will be made public. CAICT has emphasized that the introduction of OpenClaw signifies a significant milestone in the progression of artificial intelligence. Nevertheless, challenges such as the unclear delineation of boundaries in intelligent assistant agent products underscore the pressing need for standardization to offer clear guidance.