China's amended Cybersecurity Law has strengthened the consequences of security failures and widened its provisions on harmful activity overseas. The changes took effect on 1 January 2026. They affect the legal exposure of network operators and technology suppliers without replacing the law's existing territorial framework for networks inside China, according to analysis by Greenberg Traurig.1
For foreign businesses, the practical distinction is between ordinary commercial access and conduct that triggers security obligations or enforcement. The revised statute contains rules for products, network operation and critical infrastructure. Its overseas provision addresses activity endangering China's cybersecurity; additional sanctions, including asset freezes, are available where serious consequences occur.2
Product certification carries its own enforcement risk
The Chinese text requires specified critical network equipment and dedicated cybersecurity products to pass the relevant certification or testing before sale or provision. It now expressly provides penalties for supplying products that fail those requirements. This is a product-compliance route, separate from a tariff imposed at the border.2
Latham & Watkins explains that these violations can lead to confiscation of unlawful gains and fines, with business suspension or licence consequences in serious cases. Its review also highlights increased penalties for failures to perform cybersecurity duties. Maximum fines can reach RMB10 million for a business and RMB1 million for directly responsible individuals where the statutory threshold of particularly serious consequences is met. These are upper limits tied to aggravated outcomes, not standard charges for every breach.3
The firm identifies mitigation provisions as another consequential change. Cooperation, corrective action and circumstances recognised under China's administrative-penalty law can reduce liability; qualifying minor infringements that are promptly corrected may avoid a penalty. The severity and response to an incident therefore matter alongside the underlying obligation.3
The AI provision establishes a direction for further rules
The law's AI provision calls for:
加强风险监测评估和安全监管
Strengthening risk monitoring, assessment and security supervision.2
Commentary published by the Cyberspace Administration of China describes this provision as a basic framework connecting technological development, infrastructure, ethics and security. Its author, Renmin University legal scholar Ding Xiaodong, says the law provides foundations for more specific governance measures. That is an interpretation presented through the regulator's website, rather than evidence that every proposed implementation mechanism is already operating.4
The commentary also distinguishes support for training-data infrastructure from a direct statutory requirement authorising particular data use. The new provision should consequently not be read as either a general licence to use training material or a complete AI approval regime. Detailed obligations still depend on the relevant instruments and activity.4
Greenberg Traurig similarly describes the AI additions as policy and regulatory goals. Its analysis advises multinational suppliers servicing Chinese customers' IT systems from overseas to reassess their exposure, given the broader enforcement power. The firm also notes that some monetary penalties can now be imposed without first establishing failure to obey a rectification order.1
The open question for cross-border suppliers is how authorities will apply the expanded overseas provision to particular facts, and which further instruments will translate the AI policy framework into specific commercial requirements. Neither the framework language nor the increased penalty ceiling alone answers that question.1 4