China issues first AI smart glasses privacy code after secret filming outcry
Synopsis
China's CAICT has issued the world's first industry code of conduct for AI smart glasses after videos on Rokid's own forum showed users secretly filming strangers on subways, beaches, and malls — exposing a regulatory blind spot at the heart of the wearable AI boom.
Key Takeaways
China's CAICT released the industry's first code of conduct for AI smart glasses on Thursday, 26 June 2026 .
The voluntary guidelines require a 'minimum data collection' approach, active-recording indicators, and explicit user consent before filming.
Rokid , a Hangzhou -based smart glasses maker, faced public backlash after covert-recording videos appeared on its own user forum.
Footage included a flight attendant being secretly filmed while serving passengers, triggering widespread outrage.
Rokid removed the videos and banned the uploading accounts earlier in June 2026 but had not commented publicly by 27 June 2026 .
The code is currently voluntary, leaving open the question of whether binding enforcement will follow.
China has released the industry's first code of conduct for AI smart glasses, responding to widespread public anger over covert recordings made by users of the increasingly popular wearable devices. The guidelines, published on Thursday, 26 June 2026, were issued by the China Academy of Information and Communications Technology (CAICT), a research institute operating under the Ministry of Industry and Information Technology.
What the code requires
The voluntary framework calls on smart eyewear manufacturers to adopt a 'minimum data collection' approach when designing their products. Makers are also required to provide clear, visible indicators whenever cameras or microphones are active, and to obtain explicit user consent before any recording begins. While the code is non-binding, it marks the first formal industry standard of its kind in China.
Why it matters
The guidelines arrive after a surge of incidents involving smart glasses produced by Hangzhou-based tech firm Rokid. Videos circulated on Rokid's own online user forum showed members of the public being secretly filmed in everyday settings — including subway trains, parks, beaches, and shopping malls — without their knowledge. One widely shared clip depicted a flight attendant greeting passengers and serving meals, reportedly unaware she was being recorded. The footage triggered a sharp public backlash and calls for regulatory intervention.
Rokid's response
Rokid did not immediately respond to a request for comment on Friday, 27 June 2026. Earlier in June 2026, the company removed some of the offending videos from its platform and blocked the accounts responsible for uploading them. The episode has placed Rokid at the centre of a broader conversation about the responsibilities of hardware makers when their products enable covert surveillance.
The competitive backdrop
The global AI smart glasses market has accelerated rapidly, with products from multiple manufacturers blurring the line between consumer wearables and surveillance tools. China's move mirrors growing regulatory attention in other jurisdictions, where lawmakers have struggled to keep pace with the miniaturisation of cameras embedded in everyday objects. The CAICT's intervention signals that Beijing is prepared to set the terms of engagement for this category before incidents escalate further.
What's next
Industry observers will be watching whether the voluntary code is eventually codified into binding regulation, and how quickly manufacturers beyond Rokid adapt their product designs to meet the new standards. With the wearable AI market expanding, the pressure on regulators to move from guidance to enforcement is likely to intensify in the months ahead.
Point of View
Exposing a fundamental design failure in how wearable AI companies moderate user-generated content. This fits a recurring pattern in China's tech governance cycle: incidents reach public saturation, a state-linked institute issues soft guidance, and binding rules follow only if political pressure persists. The deeper question is whether 'minimum data collection' can be meaningfully enforced on a device whose core value proposition is always-on, ambient sensing — a tension no voluntary code can resolve.
NationPress
26 Jun 2026
Frequently Asked Questions
What is China's new AI smart glasses code of conduct?
China's CAICT released the industry's first voluntary code of conduct for AI smart glasses on 26 June 2026 , requiring manufacturers to minimise data collection, display clear recording indicators, and obtain explicit user consent before filming.
Why did China introduce smart glasses privacy rules?
The guidelines followed public outrage over videos posted on Rokid 's user forum showing strangers being secretly filmed in subways, parks, beaches, and malls. One widely shared clip showed a flight attendant being recorded without her knowledge.
What did Rokid do about the covert recording videos?
Rokid removed the offending videos and blocked the accounts that uploaded them earlier in June 2026 . The company had not publicly commented on the matter as of 27 June 2026 .
Is the CAICT smart glasses code legally binding?
No — the code is currently voluntary. It sets expectations for manufacturers but does not carry legal enforcement mechanisms, leaving open the possibility of future binding regulation if compliance remains patchy.
Which companies are affected by China's AI glasses privacy code?
The code applies to all smart eyewear manufacturers operating in China . Rokid , headquartered in Hangzhou , is the most prominent company named in connection with the incidents that prompted the guidelines.
