Китай впервые синхронно снял ЭЭГ у тысячи человек в разных регионах для обучения ИИ

On July 22, 2026, a Chinese scientific team unveiled a new device for recording electroencephalograms (EEG) and, for the first time in the world, synchronously recorded the brain signals of more than a thousand people located in different regions of the country. This was reported by the CCTV Finance channel (央视财经), and the news was distributed by the business outlet 36Kr.

What the breakthrough is about

For the first time in the world, it became possible to simultaneously and synchronously record EEG from more than a thousand people spread out geographically — previously, large-scale collection of brain signals remained a laboratory experiment involving single-digit or double-digit numbers of participants. According to the developers, this is a key step toward training large neural models and creating universal brain-computer interface (BCI) technologies.

CCTV Finance calls the brain-computer interface one of the most important industries of the future, and large-scale, high-quality EEG data the foundational basis for moving the technology from the laboratory into industry.

• Date — July 22, 2026, the presentation was shown in the afternoon
• Scale — synchronous EEG collection from more than 1000 people at the same time
• Geography — participants were located in different regions, not in a single laboratory
• World record — the world's first cross-regional synchronous collection of this scale
• Goal — accumulating data to train neural foundation models

What problems had to be solved

The development team named two main technical difficulties. The first was preserving recording accuracy while miniaturizing the device: the device had to be made smaller without losing the quality of the EEG signal. The second was overcoming the effect of network latency and achieving millisecond-level time synchronization among a multitude of devices in different regions.

Millisecond-level time alignment accuracy is essential here. Without it, signals from different people and devices cannot be analyzed within a single coordinate system, meaning it would be impossible to combine a thousand separate recordings into one coherent array. It is precisely synchronization that turns scattered EEG streams into a dataset suitable for training models.

Why this matters for artificial intelligence

The collected data — synchronous EEG recordings from more than a thousand people — will be used to train neural foundation models. As CCTV Finance explains, the large-scale EEG collection technology will gradually accumulate such data and use it to train models.

Direct collection of neural signals changes the very nature of the training material for AI. Until now, models learned from indirect information — text, images, video. Recording brain activity allows AI to read a person's cognitive state directly, rather than through intermediary media.

Training material for AI is no longer limited to text, images, and video — now, through neural signals, it is possible to directly understand a person's cognitive state. — from a CCTV

Finance (央视财经) report

What this means

Scaling up EEG collection is primarily about data, not about a single gadget: for brain interfaces to move from the laboratory into industry, large and synchronized datasets of brain signals are needed. The Chinese team showed that such collection is technically possible at once across thousands of people in different cities, and declared this a world first. Neither the name of the developing institute or company, nor the exact parameters of the device — the number of channels, the recording frequency — are disclosed in the CCTV Finance report, so the methodology cannot yet be independently assessed.

Frequently asked questions

How many people took part in the synchronous EEG collection?

According to CCTV Finance, brain signals were synchronously recorded from more than a thousand people at the same time. This is the world's first case of cross-regional EEG collection on this scale.

Why collect EEG from thousands of people at once?

Large-scale, high-quality EEG data is needed to train neural foundation models. According to the developers, this will allow AI to understand a person's cognitive state directly from neural signals, not only from text, images, and video.