In an era where facial recognition technology is increasingly embedded in daily life, from unlocking smartphones to streamlining payments, China has introduced robust regulations to ensure its responsible use. On March 21, 2025, the Cyberspace Administration of China (CAC) and the Ministry of Public Security (MPS) released the Security Management Measures for the Application of Facial Recognition Technology, effective from June 1, 2025. These measures, supplemented by a clarifying notice from the CAC on March 30, 2025, mandate registration for companies processing facial data of over 100,000 individuals and provide a clear framework for compliance. For British businesses operating in or entering the Chinese market, understanding and adhering to these rules is essential to safeguard operations and protect personal data.

The New Regulatory Landscape

China’s facial recognition regulations are part of a broader effort to strengthen data protection under the Personal Information Protection Law (PIPL), enacted in 2021. The Security Management Measures aim to balance innovation with the protection of individual privacy, addressing concerns about the misuse of sensitive biometric data. The rules apply to any organisation, domestic or foreign, processing facial recognition data in China, with a particular focus on those handling large datasets. According to the CAC, companies storing facial data of more than 100,000 individuals must register with their provincial-level cyberspace administration within 30 working days of reaching this threshold.

Recognising the compliance burden, the CAC introduced a grace period for companies that hit this threshold before June 1, 2025, allowing them until July 14, 2025, to complete registration. This transitional measure reflects China’s pragmatic approach to implementation, ensuring businesses have time to adapt without immediate disruption. Additionally, the CAC issued detailed ‘Instructions for Filling in the Facial Recognition Technology Application Filing System (First Edition)’, accessible via the Personal Information Protection Business System or the National Cyberspace Administration Government Affairs Hall on the CAC’s website. These guidelines outline the registration process, required documentation, and compliance expectations, making it easier for companies to navigate the system.

Why Compliance Matters

Facial recognition technology is widely used in China across sectors like retail, finance and security, but its rapid adoption has raised concerns about privacy and data security. China’s facial recognition market is projected to reach £7.2 billion by 2027, driven by applications in smart cities and public safety. However, high-profile cases, such as the 2021 fine imposed on a Hangzhou zoo for collecting facial data without consent, underscore the risks of non-compliance. The zoo was ordered to delete the data and issue a public apology, highlighting China’s growing emphasis on enforcement.

For British businesses, compliance is not just about avoiding penalties; it’s about building trust in a market where data protection is increasingly scrutinised. Robust cybersecurity measures, including compliance with data laws, are critical for protecting investments in China. Failure to register or properly handle facial data could result in fines, operational restrictions, or reputational damage, particularly for companies in sectors like technology, retail, or hospitality that rely on facial recognition for customer engagement.

How to Register: A Step-by-Step Guide

The registration process is designed to be straightforward, with all steps completed online via the Personal Information Protection Business System. Companies must first create an account on the platform before uploading the required documents, which include:

• A Basic Information Form of Personal Information Processor, detailing the company’s operations and data processing activities.
• A Facial Recognition Technology Application Record Form, outlining the scope and purpose of facial data use.
• A Personal Information Protection Impact Assessment (PIPIA), assessing the legality, necessity, and risks of data processing.
• Scanned copies of the Unified Social Credit Code Certificate, legal representative’s ID, agent’s ID, Power of Attorney, and Letter of Commitment, all stamped with the company’s official seal.

The CAC reviews submissions within 15 working days, updating the application status to “Filing Completed,” “Returned for Improvement,” or “Review Failed.” If supplementary materials are required, companies have 10 working days to provide them, or the process is terminated. The CBBC advises seeking professional support, such as from its Information Systems team, to ensure compliance with China’s data laws and to localise global systems effectively.

Conducting a Personal Information Protection Impact Assessment (PIPIA)

A cornerstone of the new regulations is the requirement to conduct a PIPIA, as mandated by the PIPL. This assessment evaluates the legality, legitimacy, and necessity of facial data processing, alongside the potential impact on individual rights and the effectiveness of protective measures. The Filing Instructions provide a tailored template for facial recognition, requiring companies to disclose technical specifications, data collection and storage methods, standard operating procedures, and the ethical basis for data use. For example, companies must clarify whether facial data is used for automated decision-making, such as targeted advertising, and detail the infrastructure and technology providers involved.

The PIPIA process encourages transparency and accountability, aligning with international best practices. The PIPIA requirement has driven companies to adopt more robust data governance frameworks, enhancing trust among consumers and regulators alike.

Easing the Transition

The CAC’s notice reflects a pragmatic approach to regulation, balancing enforcement with flexibility. The grace period for pre-June 2025 data processors and the detailed Filing Instructions demonstrate China’s commitment to supporting businesses during this transition. For British companies, this is an opportunity to align with China’s evolving data protection regime while leveraging tools like the CBBC’s Business Guides, which offer insights into regulatory compliance and market navigation.

The post What Are China’s New Facial Recognition Regulations? appeared first on Focus - China Britain Business Council.

China and the UK both have ambitions to become global leaders in innovation, and Artificial Intelligence (AI) is helping achieve these goals. Adoption of AI is increasingly seen in a wide variety of sectors and the education sector is no exception.

A rise in e-learning, app-based education and Massive Open Online Courses, or MOOCs, has led to the digitisation of educational content; and AI helps to personalise learning for students, based on their needs and progression.

By harnessing ‘Natural Language Processing’ or speech recognition, and combining this with smart technology developments such as facial and voice recognition, AI can detect micro-expressions in student miscomprehension and faults in pronunciation. This, in the instruction of English language, could be revolutionary.

With the rise of household incomes and the introduction of the two-child policy in China, demand for English language learning is growing. Increasingly parents are introducing English language classes to their children at pre-school ages.

“[AI] can really decrease the pressure on the problems of education resource shortage,” says Ying Weihai, Professor at Shanghai Jiaotong University and Founder of Jiangsu Kunhui Biotechnology Corporation. “While there may be weaknesses in AI in China, such as in basic theory, there is great strength in application, such as with driverless cars by Baidu, medical imaging and genetic data by Tencent, and voice recognition by Keda Xunfei.”

 It is developing facial and voice recognition to perceive student comprehension in classrooms

It is, therefore, no surprise that some of China’s leading, existing English language providers have begun to utilise AI in their programmes. Liulishuo is an app-based AI English teacher that provides a personalised and adaptive path for each of its learners. And Hujiang provides remote learning to rural schools through its online platform CCTalk. It is also developing facial and voice recognition to perceive student comprehension in classrooms in order to improve the student/teacher interaction.

There are a number of apps developing facial and voice recognition to perceive student comprehension in classrooms in order to improve the student- teacher interaction.

There are many other Chinese digital English language competitors, such as: VIPkid, 51talk, ITutor Group, and TAL Education Group, and they too, will need to look at developing AI applications to stay competitive.

The UK has strength in its depth of high-quality academic content produced by publishers such as Oxford University Press, Cambridge University Press and Pearson, who recently sold their China-based learning centres Wall Street English in order to focus on content production. It has partnered with Microsoft Research Asia to launch ‘Longman English +’, an English language learning application that integrates both Pearson’s ‘Longman Welcome to English Curriculum’ and Microsoft’s Artificial Intelligence developments. Pearson’s actions reflect just one example of how UK firms could access the highly competitive ‘EdTech’ market by providing high-quality content to existing tech firms.

The UK’s EdTech industry is one of its fastest growing sectors, accounting for four percent of the digital industry – the same amount as Financial Technology or ‘FinTech’ firms. Many EdTech start-ups are introducing AI in their services, such as Tutorean, an app that uses AI to improve tutor/student matches or Zzish, whose virtual teaching assistant identifies struggling students, diagnoses learning gaps and then recommends resources. Magpie Education also uses AI to automatically assess and monitor the progress of students in order to recommend learning pathways and provide personalised content.

With the UK’s global reputation for education and innovation, and China’s leadership in scaling technology, there are huge opportunities for partnerships

Such innovative UK AI start-ups could also offer their services, particularly in the field of English language, to both offline and online education providers, in order to help such firms keep up with the quickly evolving education market. With the UK’s global reputation for education and innovation and China’s leadership in scaling technology, there are huge opportunities for partnerships.

“For China to achieve its ambition of becoming a leader in AI and smart manufacturing, it will need to welcome partnerships with foreign companies and universities – at least in the short term,” says Nathan Lawes, consultant at IoT ONE. “Despite taking a protectionist stance in the past, China has committed to opening up its markets now that domestic players have become national champions. If they wish to become global champions, it makes sense to open up the industry further in order to compete (and collaborate) with their European and American counterparts”.

The post Harnessing AI for English language instruction appeared first on Focus - China Britain Business Council.

Facial-recognition technology is central to the current Artificial Intelligence (AI) boom taking place in China. In recent months, three of the country’s vision-based solution start-ups have attained ‘unicorn’ status, with market valuation of at least USD $1 billion. No other country can claim to have a single one. The tech giants Baidu, Alibaba and Tencent (collectively known as BAT), are also heavily promoting facial recognition software on their respective platforms.

Government support has undoubtedly helped, with last year’s State Council proposal underlining plans to transform the market into a US $150 billion industry by 2030.[1] The technology is also already being applied to the country’s extensive internal security network, further increasing its value to Beijing. However, it is in the retail, payment and e-commerce sectors that the technology’s potential may reap the most rewards.

One of several major companies leading the Chinese facial-recognition movement is SenseTime. The firm’s innovative computer vision and deep-learning technology software has led to over 400 partnerships with domestic and foreign companies from sectors as wide-ranging as surveillance, autonomous vehicles and retail, whilst attracting sufficient investment to ensure its status as the world’s highest-valued AI start-up.

The application of SenseTime’s software in brick-and-mortar retailers such as Suning, one of China’s largest electronics providers, has helped develop what is being labelled ‘Smart Retail’. Facial-scanning software can efficiently gather data related to age, gender, and mood from in-store shoppers, helping the retailer adapt their marketing and customer service strategies accordingly. “By providing our software, we can help retailers develop their own applications,” a company spokesperson told FOCUS.

The tech has also allowed non-traditional players to enter the ‘Smart Retail’ market by creating unstaffed, cashless stores. In-store at one of BingoBox’s 158 Chinese outlets, items are identified via cameras utilising sophisticated image recognition technology and purchased using the customer’s Alipay or WePay app. Similar methods have also been employed by Tencent, which launched its own unstaffed and cashless store last year, as well as start-ups such as TakeGo and Fxbox.

Alibaba has gone one step further. Having already entered the market with 2016’s ‘Tao Café’, the tech giant last September collaborated with KFC by introducing a ‘smile to pay’ service. Integrated within the company’s Alipay payment app, customers are scanned and required to smile to confirm their payment.

KFC’s smile to pay service, required customers to smile to confirm their payment

Elsewhere, Shenzhen-based start-up Malong Technologies has found great success applying deep-learning AI technology to product analysis. By processing petabytes of data, the equivalent to the entire photo catalogue currently stored on Facebook, Malong’s AI has achieved human-level analytical precision that can be applied to the retail and e-commerce sectors as well as in manufacturing and security.

For e-commerce giants such as Alibaba and JD.com, the technology provides much-needed additional protection for consumers against fraud and counterfeit. By analysing product images, AI bots are able to spot cheap knock-offs and fakes and notify both the customer and platform provider before any damage is done.

Not satisfied with already possessing the world’s most developed e-commerce environment, as well as a rapidly growing retail industry, innovation provided by AI will further boost China’s economic standing. The impact is already evident with Alibaba’s record-making US $25 billion ‘Singles Day’ sales being attributed to the use of AI in customer service and marketing. It is therefore perhaps unsurprising that a recent PwC report estimated that, by 2030, AI could boost Chinese GDP by up to 26 percent.

Opportunities within the market are not limited or restricted solely to Chinese firms, and, with a burgeoning AI industry, UK companies may also stand to benefit. For Mark Hedley, CBBC’s Senior Director for ICT, Britain is in position to become a key partner in AI with China: “The UK’s recently announced AI Sector Deal looks set to sustain the growth of the sector. This links directly with China’s own plans to use AI technologies as a key growth driver, creating new opportunities for collaboration between the two countries”.

This readiness for cooperation is reflected by the upcoming UK mission to Guiyang that, for Hedley, “represents the best in British innovation, including healthcare AI, machine learning, video analytics, cybersecurity and more. Building on 2017’s UK-China Big Data Collaboration, the delegation will help cement UK-China ties in innovation”.

However, while UK firms such as Deepmind, Babylon and Benevelont.AI have already had a major impact on the global AI scene benefitting from the scientific resources Britain possesses,[9] China has its own unique attraction for foreign start-ups. This is a point raised by David Bian, the Strategic Alliance Manager at the UK multinational company ARM’s Shenzhen base. “While innovation and R&D is strong in the UK, China provides the opportunity for application,” Bian states. “In addition, the Shenzhen government is very efficient when it comes to innovation. Things are being updated all the time, faster than in the UK.”

ARM’s own Accelerator and AI Ecosystem Consortium (AIEC), have provided one such opportunity for growth by providing both Chinese and foreign AI start-ups with resources. BAT, in their position as ‘national champions’ have also established platforms that promote cooperation and competition in the field and access to the companies’ massive pools of data.

UK companies may also be in the advantageous position to provide what some Chinese facial and vocal recognition companies lack – a variety of data. “Current drawbacks for us include the lack of difference between faces and languages occurring in Chinese realtime data,” a SenseTime spokesperson stated. “There definitely lies an opportunity for UK companies to collaborate to improve our respective capabilities.”

While innovation and R&D is strong in the UK, China provides the opportunity for application

However, the AI industry carries with it perhaps more concern over Intellectual Property (IP) leakage than almost any other, with Chinese companies especially cautious. Any UK company seeking Joint Venture or market entry opportunities must take this into account.

As facial recognition technology becomes increasingly advanced, worries over data protection have grown, with recent events at Facebook and Cambridge Analytica emphasising concerns. While China has established previously non-existent Data Protection and Cybersecurity Laws, they are, at best, a work in progress, with major issues related to clarity and regulatory responsibility threatening to put prospective UK partners off from entering the market.

The definition of ‘personal information (sensitive)’ and ‘important data’ in this context is particularly troublesome and something Bian alludes to: “BAT [and others] are only collecting unsensitive data, although what constitutes sensitive and unsensitive may be slightly different in China and abroad”.

This remains the key concern, and the schism partitioning data attitudes may be set to widen as the EU’s General Data Protection Regulation (GDPR), coming into force in May of this year, explicitly labels biometric data, the key source of China’s ‘Smart Retail’ revolution, ‘sensitive’ in the realm of retail and e-commerce. This may have a direct influence on how AI technology develops in China and the West.

China’s impressive growth has been built on the back of detailed roadmaps designed to deliver commercial success and AI is no exception. Supported by huge pools of biometric data, the major BAT tech firms have created a platform and network to innovate the already well-developed retail and e-commerce industries. While concerns surrounding data protection and privacy rights remain, Chinese enthusiasm for vision-based AI solutions threatens to leave the West behind. UK tech companies, backed by ambitious new government plans, would provide the perfect partner with which to take the next step.

The post Facial recognition software is used across China’s retail sector but what about privacy laws? appeared first on Focus - China Britain Business Council.