China Releases First AI-Plus-BCI Medical Device Standard

China has published the world’s first product standard for a brain-computer interface (BCI) medical device that uses artificial intelligence to process brainwave data, a regulatory milestone for a technology that sits at the intersection of neuroscience and machine learning. The National Medical Products Administration (NMPA) approved and released the standard on 14 September 2026, according to Eastmoney, citing CCTV News.

Designated YY/T 2029-2026, the standard sets quality requirements and evaluation methods for the electroencephalogram (EEG) datasets used to train AI algorithms in BCI medical devices. It takes effect on 1 September 2027.

What the Standard Covers

The full title of the standard translates to “Medical devices utilizing brain-computer interface technology — Quality requirements and evaluation methods for EEG datasets used for artificial intelligence algorithms.” According to China’s official NMPA announcement, the standard was reviewed and approved before publication.

The measure applies to datasets used in the research, development, production, and quality-control stages of BCI medical devices in which AI algorithms post-process EEG signals. As Xinhua reported via Quanzhou Web, it systematically standardizes technical requirements and test methods across the entire data pipeline — collection, processing, annotation, storage, and access — to help companies resolve common industry problems such as inconsistent dataset quality-control benchmarks and non-standardized dataset construction.

Why It Matters

The NMPA framed the standard as essential to both treatment efficacy and patient safety. “A BCI medical device is like an intelligent machine that can ‘read’ brain signals,” the agency said, as reported by CCTV News. “It needs to ‘learn’ from large amounts of EEG data and ‘train’ AI algorithms in order to accurately understand the patient’s intentions. This affects not only treatment efficacy but is also closely tied to patient safety risks.”

The standard provides what the NMPA called a unified “quality ruler” for EEG datasets, which it says reaches an internationally leading level. To support major BCI innovation and meet urgent industry demand, the regulator used an expedited procedure to initiate and approve the standard.

This is China’s third BCI medical device standard, following the terminology standard YY/T 1987-2025, which took effect on 1 January 2026, and a second standard published in the interim, per Shanghai’s Medical Products Administration. Because it specifically governs AI-based EEG data processing, it is described as the first such product standard in the world.

A Fast-Growing Industry

The standard arrives amid a broader national push to commercialize BCI technology. China’s BCI market was valued at RMB 3.20 billion (about 320 million yuan) in 2024 and is projected to reach RMB 6.14 billion by 2028, according to public data cited by Rednet, republishing an Economic Daily report.

Seven ministries, including the Ministry of Industry and Information Technology, have issued implementation guidelines aimed at cultivating two to three globally influential BCI enterprises and building a competitive industrial ecosystem by 2030. Anteil, chief engineer at the China Academy of Information and Communications Technology, described the policy as a key step for China to move from “follower” to “parallel runner” and potentially “leader” in the field.

BCI technology establishes an information channel between the brain and a machine, enabling interaction between biological and machine intelligence. Its primary applications are in healthcare — diagnosing and rehabilitating conditions such as epilepsy, Parkinson’s disease, and Alzheimer’s disease — as well as industrial manufacturing and consumer products. Several Chinese companies have already obtained medical device registration certificates for BCI motor-rehabilitation equipment used in stroke recovery.

Challenges Ahead

Industry experts caution that significant obstacles remain. Anteil noted that EEG signal precision and acquisition stability, neural decoding algorithms, and the biocompatibility of implantable electrodes all require further improvement. He also pointed to an immature industrial ecosystem and unresolved ethical and privacy concerns around brain data.

Yu Yuntao, technical director at the China Electronics Standardization Institute, said individual agencies should develop technical, product, service, and testing standards, and actively participate in international standard-setting to advance Chinese standards globally.

What to Watch

The new standard’s practical impact will emerge only after it takes effect on 1 September 2027, when manufacturers of AI-powered BCI devices will need to demonstrate that their training datasets meet the new quality benchmarks. For now, the measure signals China’s intent to shape the rules of a field it hopes to lead — and to do so before the technology reaches broad clinical adoption.