Time
Click Count
On May 8, 2026, China’s Ministry of Industry and Information Technology (MIIT) and other authorities jointly released the national standard Grading of Artificial Intelligence Terminal Intelligence (GB/Z 177–2026). The standard introduces mandatory evaluation requirements for biometric authentication—including 3D facial recognition, iris/vein identification—and cloud security gateways under its newly defined ‘Category N’ classification. It takes effect immediately and directly affects export compliance pathways for intelligent security hardware destined for the EU, Southeast Asia, and the Middle East—particularly regarding alignment with CE, UKCA, and ID-SNI certification frameworks. Hardware manufacturers, exporters, and certification service providers operating in AI-powered access control, identity verification, and edge-cloud security should closely monitor implementation implications.
On May 8, 2026, the People’s Republic of China officially published GB/Z 177–2026, titled Grading of Artificial Intelligence Terminal Intelligence. This is the first national standard to formally define and mandate technical evaluation criteria for AI-enabled terminal devices featuring advanced biometric modalities (e.g., 3D facial recognition, iris recognition, vein pattern recognition) and cloud-integrated security gateways. The standard classifies such technologies under ‘Category N’ and requires conformity assessment prior to market entry or export. It entered into force on the date of publication.
Exporters shipping AI terminals—including smart door locks, access controllers, and identity verification kiosks—to the EU, Southeast Asia, and the Middle East are directly affected because the standard now serves as a prerequisite for obtaining or maintaining CE, UKCA, and ID-SNI certifications. Compliance with GB/Z 177–2026 may become a formal condition for mutual recognition agreements between Chinese certification bodies and foreign accreditation authorities.
OEMs and ODMs integrating 3D facial, iris, or vein recognition modules—or embedding cloud security gateway functionality—into end products must now align their design, testing, and documentation processes with Category N requirements. This includes updated test protocols for liveness detection, template protection, and secure cloud handshaking mechanisms.
Domestic and international certification organizations accredited to issue CE, UKCA, or ID-SNI marks may need to update their assessment checklists and audit procedures to reflect GB/Z 177–2026’s Category N criteria—especially where local regulatory authorities reference the standard in official guidance or technical annexes.
While GB/Z 177–2026 is effective immediately, detailed technical implementation guidelines, testing methodologies, and certification roadmaps have not yet been publicly issued. Companies should track announcements from MIIT, the Standardization Administration of China (SAC), and the China National Accreditation Service for Conformity Assessment (CNAS).
Manufacturers and exporters should map existing product portfolios against the standard’s definition of Category N: specifically, devices incorporating 3D facial recognition, iris/vein biometrics, or cloud security gateway functions. Products meeting any of these criteria require reassessment—even if previously certified under legacy frameworks.
The standard’s immediate effectiveness signals regulatory intent but does not automatically invalidate prior certifications. However, new applications or major product revisions submitted after May 8, 2026, are expected to undergo Category N evaluation. Companies should verify whether their current certification scope already covers the relevant sub-clauses before assuming continuity.
Procurement teams should confirm component-level compliance (e.g., biometric sensor modules, secure boot firmware); R&D and QA teams should revise test plans; and regulatory affairs teams should prepare updated technical files—including risk assessments for spoofing resistance and cloud data integrity—aligned with GB/Z 177–2026 Annexes.
Observably, GB/Z 177–2026 functions less as a standalone technical specification and more as a strategic anchor point for China’s broader AI governance framework—linking domestic standardization to international market access. Analysis shows that its inclusion of cloud security gateways and multi-modal biometrics reflects an emerging emphasis on system-level trustworthiness, not just algorithmic accuracy. From an industry perspective, this standard is best understood not as a final compliance endpoint, but as the first formalized layer in an evolving regulatory stack governing AI at the edge. Its real-world impact will depend heavily on how foreign regulators incorporate it into mutual recognition dialogues—and whether third-party labs rapidly develop validated test methods for Category N claims.

This standard represents a structural shift: it transforms voluntary technical differentiation (e.g., ‘AI-enhanced lock’) into a codified, enforceable classification. For global supply chains, it adds a new upstream checkpoint—not only for hardware design, but for interoperability assumptions between on-device AI and cloud infrastructure. Current implementation remains transitional; sustained monitoring of SAC-issued interpretations and CNAS-accredited lab capabilities is essential.
Information Sources: Official release notice by the Standardization Administration of China (SAC), May 8, 2026; Joint announcement by MIIT, SAC, and the State Administration for Market Regulation (SAMR); GB/Z 177–2026 full text (published version). Note: Technical implementation guidelines, test method specifications, and official translations remain pending and are subject to ongoing observation.
Recommended News