China Algorithm Filing Landscape: 2026 Mid-Year Report

Three and a half years after China’s algorithm filing system launched in March 2022, the landscape has evolved dramatically. What began as a relatively straightforward registration requirement has expanded into a multi-layered regulatory framework covering algorithmic recommendations, deep synthesis, generative AI, and on-device AI — with more than 700 algorithms filed in the generative AI category alone.

This report synthesises data from the Cyberspace Administration of China (CAC) public filing announcements through June 2026, plus proprietary data from client engagements and industry sources, to paint a comprehensive picture of where algorithm filing stands today and where it is heading.

Headline Numbers

China Algorithm Filing Landscape: 2026 Mid-Year Report - AI Law & Regulation infographic

As of June 2026, the CAC has announced 26 batches of algorithm filing registrations. The cumulative numbers tell a story of steady growth and expanding scope.

MetricValue (as of June 2026)
Total algorithm filings (all categories)670+
Generative AI service registrations190+
Deep synthesis algorithm filings240+
Algorithmic recommendation filings210+
Companies with filed algorithms~450
Foreign-invested company filings~28
Average filing processing time37 days
Batch announcement frequency~Every 3-4 weeks

These numbers include only algorithms that have received formal approval and been publicly announced. There are typically 30-50 additional filings in process at any given time, at various stages of review.

Growth Trajectory

The algorithm filing system has gone through three phases so far.

Phase 1 (March 2022 – January 2023): The initial rollout. The first nine batches focused primarily on algorithmic recommendation systems at major internet platforms. Filings were concentrated among the largest domestic tech companies, and the process was still being refined. Total filings in this period: approximately 80 algorithms.

Phase 2 (February 2023 – December 2024): Generative AI boom. The launch of generative AI regulations in late 2022 and early 2023 triggered a wave of new filings as Chinese companies rushed to register their large language model services. The filing process matured, with more standardised requirements and faster review times. Total filings in this period: approximately 280 algorithms.

Phase 3 (January 2025 – June 2026): Diversification and deepening. In the most recent phase, growth has come from multiple directions: more deep synthesis filings (driven by video generation and virtual human applications), more on-device AI filings (as phone manufacturers add on-device generative AI features), and a steady stream of new generative AI services. Foreign company participation has also increased, though from a low base. Total filings in this period: approximately 310 algorithms in just 18 months — more than the first two phases combined.

Category Breakdown

The filing landscape is roughly evenly divided across three major categories, with each category showing different growth dynamics.

CategoryEst. Total FilingsShareGrowth Rate (YoY)
Algorithmic recommendation~21031%+18%
Deep synthesis~24036%+62%
Generative AI services~19028%+45%
Personalised push / other~305%+12%
Total~670100%+38%

Deep synthesis has been the fastest-growing category over the past year, driven by the explosion of AI video generation tools, virtual human/avatar services, and voice cloning applications. The deep synthesis label requirement — which mandates that AI-generated content be clearly labelled — has also pushed more companies to formalise their filings.

Generative AI services continue to grow steadily as more companies launch LLM-based products and as existing services add features that trigger additional filing requirements. The category is also seeing more specialised applications beyond general-purpose chatbots — coding assistants, legal AI, medical AI, and education tools are all growing.

Algorithmic recommendation is the most mature category, with growth coming from new entrants rather than from major platforms (which largely completed their initial filings in 2022-2023).

Foreign Company Participation

Foreign-invested companies still represent a small fraction of total filings, but the number is growing. As of mid-2026, we estimate that approximately 28 algorithms filed with the CAC are from foreign-invested entities — about 4% of the total.

YearForeign-Invested FilingsCumulative
2022~22
2023~68
2024~917
2025~724
2026 (H1)~428

The growth in foreign company filings reflects several trends: more international AI companies establishing China entities, the normalisation of the filing process (making it more predictable for foreign companies), and the increasing importance of the Chinese market for AI products globally.

For foreign companies, the filing journey typically looks different from domestic companies. They often need more time for preparation — including entity setup if they do not yet have a China entity, translation and documentation work, and internal alignment across global teams. The actual filing review time, however, is generally the same for foreign-invested and domestic companies.

Industry Distribution

Which industries are driving algorithm filing growth? Based on our analysis of filed algorithms and their service descriptions, the industry mix has been shifting.

Industry / SectorShare of FilingsTrend
General internet platforms28%Stable
Generative AI / LLM services22%Growing rapidly
E-commerce & retail12%Growing steadily
Media & entertainment10%Growing (video/audio generation)
Education & training7%Growing
Enterprise software / SaaS6%Growing
Finance & fintech5%Stable
Healthcare & medical AI4%Growing
Manufacturing & industrial3%Emerging
Automotive / smart mobility3%Growing

The decline in general internet platforms’ share is not because they are filing less — it is because other categories are growing faster. The platform companies were early filers and have largely completed their initial portfolios; new filings from them tend to be incremental (new features, new products, updates).

Processing Times

One of the most common questions from companies considering filing is: how long does it take? The answer has been getting better.

PeriodAverage Processing TimeRange
202255 days35-90 days
202344 days25-75 days
202439 days20-65 days
202537 days18-60 days
2026 (H1)35 days15-58 days

The average processing time has dropped by 36% since 2022, as the CAC has gained experience, standardised its review processes, and built out its review capacity.

That said, the average masks significant variation. Straightforward filings in well-understood categories can be approved in as little as 15-20 days. Complex filings — especially new types of generative AI services or applications that touch on sensitive areas — can still take 60 days or more, often with multiple rounds of questions.

The number of rounds of regulator questions also varies. Our data shows:

Number of Review RoundsShare of Filings
1 round (direct approval)~25%
2 rounds~40%
3 rounds~25%
4+ rounds~10%

Only about a quarter of filings are approved in a single round. The majority go through at least two rounds of questions and responses. This is why preparation is so important — the more complete and well-structured your initial filing, the fewer rounds you will likely need, and the faster you will get approved.

Common Reasons for Delays

Based on our experience and industry data, these are the most common reasons filings take longer than average.

  1. Incomplete or unclear algorithm descriptions. Vague or overly technical descriptions that do not clearly explain what the algorithm does and how it works generate more questions. The regulator needs to understand your service, and if your filing is too technical or too abstract, they will ask for clarification.

  2. Training data sourcing questions. For generative AI services, questions about training data sources, copyright compliance, and data licensing are among the most common second-round questions. Be prepared to describe your data sources clearly and explain how you ensure compliance.

  3. Safety mechanism effectiveness. Regulators want to see that your content moderation and safety systems are actually effective, not just theoretical. They may ask for testing data, real-world effectiveness metrics, or details about your human-in-the-loop systems.

  4. User protection measures. Questions about user rights, transparency features, appeal mechanisms, and minor protection are common. The more user-facing your product is, the more scrutiny these areas receive.

  5. Scope classification issues. Sometimes there is disagreement about which filing category a product falls into, or whether a particular feature triggers additional filing requirements. Resolving these classification questions can take extra time.

What to Expect in H2 2026 and 2027

Based on the current trajectory and signals from regulators, we expect several developments in the second half of 2026 and into 2027.

Faster processing, but more scrutiny on quality. Average processing times will likely continue to improve modestly, but the bar for what counts as a complete and compliant filing is also rising. Regulators are becoming more sophisticated in their review, and companies that submit incomplete or low-quality filings will face more pushback.

More specific guidance for AI agents. AI agent technology is advancing rapidly, and the regulatory framework is still catching up. We expect more specific guidance on how agentic AI services fit into the existing filing framework — including questions about autonomous decision-making, tool use, and accountability for agent actions.

Deepening of enforcement. Enforcement activity has been increasing steadily, and this trend will continue. Regulators are focusing not just on whether companies have filed, but on whether they are maintaining their filings properly, fulfilling ongoing obligations, and complying with requirements like deep synthesis labelling.

More cross-border coordination. As AI regulation develops globally, there is growing interest in cross-border regulatory cooperation. While full harmonisation is unlikely, we may see more information sharing and coordination between Chinese regulators and their counterparts in other jurisdictions — especially on issues like AI safety and foundation model governance.

On-device AI clarification. The regulatory framework for on-device generative AI is still evolving. We expect more detailed guidance on the specific obligations for on-device AI services, including how filing works when the model runs on the user’s device rather than in the cloud.

Key Takeaways for Foreign Companies

  1. Filing is manageable — but start early. The process is more predictable and faster than many foreign companies expect, but you still need to budget 6-10 weeks from start to finish for a typical case. Starting early gives you buffer for unexpected questions.

  2. Preparation quality drives speed and outcome. Companies that invest time in preparing a thorough, well-structured filing package — with clear descriptions, complete documentation, and well-documented safety measures — tend to get approved faster and with fewer rounds of questions.

  3. Filing is not a one-time project. It is an ongoing obligation. You need to update your filing when you make material changes, review it annually, and stay current with regulatory developments. Build compliance into your product development process, not just the launch process.

  4. The bar is rising. As the filing system matures, expectations are increasing. What was acceptable in 2023 or 2024 may not be sufficient in 2026-2027. If you filed early, it is worth reviewing your filing against current standards to make sure it is still compliant.

  5. Foreign companies can and do file successfully. While foreign-invested companies still represent a small share of total filings, the number is growing, and the process works. The key is understanding the requirements, working with experienced advisors, and giving yourself enough time.


Data in this report is based on public CAC announcements through June 2026, supplemented by proprietary client data and industry sources. Filing counts are estimates based on publicly available information and may not include all filed algorithms. This report is for informational purposes only and does not constitute legal advice.

How AI Regulation Maps to Your China Entry Timeline

Foreign companies often ask when in their China entry process they should start planning for AI regulation. The short answer: before you begin product localisation. Algorithm filing is not a last-minute compliance checkbox — it requires scope analysis, document preparation, and iterative feedback with the Cyberspace Administration of China (CAC) that typically takes four to eight weeks. If your launch timeline is tight, you should start the filing assessment during the product design phase, not after you have a working demo.

The regulatory landscape also interacts with your entity structure. A wholly foreign-owned enterprise (WFOE) that provides AI services directly to Chinese users has different filing obligations than a representative office or a cross-border service model. Getting the entity structure right first can simplify — or complicate — every subsequent compliance step.

For generative AI specifically, the registration process has three gates: pre-launch security assessment, formal filing, and ongoing supervision. Companies that plan for all three from day one tend to launch faster than those that treat filing as a single, isolated task.

How to Build an In-House Compliance Team in China

As your China operations grow, you will eventually need to bring some compliance functions in-house rather than relying entirely on external advisors. The question is: when, and what roles do you need?

For early-stage companies with a small China team, it usually makes sense to outsource most compliance work to external counsel and consultants. As you grow to 20-30 China employees and have a launched product, you should consider hiring your first compliance person — typically a senior compliance manager or data protection officer who can handle day-to-day compliance coordination and work with external advisors on more complex issues.

As you scale further — say 50+ employees and multiple products — you may need a small team: a compliance lead, someone focused on data protection, and someone focused on regulatory filing and government relations. The exact structure depends on your industry, your regulatory risk profile, and your growth trajectory. The key is to grow the team proactively — before a crisis or a regulatory gap forces you to — but not so fast that you are paying for capacity you do not need.

Annual Compliance Planning for AI Companies in China

AI regulation in China is not a one-time project — it is an ongoing obligation. Companies that treat it as a checkbox exercise consistently run into problems. Those that build structured annual compliance programmes, on the other hand, find that the burden decreases over time as systems mature.

A well-structured annual compliance calendar should include: quarterly reviews of algorithm changes to identify filing update triggers, an annual filing review and renewal process, periodic deep synthesis label audits, regular training for product and engineering teams, an annual data protection impact assessment for high-risk processing, and monitoring of regulatory developments with impact assessments.

The rhythm matters. Quarterly check-ins keep compliance on the radar without overwhelming the team. Annual reviews provide an opportunity to step back and assess whether the overall programme is still fit for purpose. And regular training ensures that new team members understand their compliance responsibilities and that existing knowledge stays current as regulations evolve.

The Difference Between Filing, Registration and Security Assessment

One of the most common points of confusion for foreign companies is the relationship between algorithm filing (算法备案), generative AI service registration (生成式人工智能服务备案), and security assessment (安全评估). These are three distinct processes with different triggers, timelines and regulators.

Algorithm filing is the broadest requirement: it applies to any algorithmic recommendation system, deep synthesis service, or personalised push system with public-facing users. Generative AI service registration is a subset — it applies specifically to large language model services and requires additional steps, including a pre-launch security assessment. The security assessment itself is the deepest review, examining training data sources, content generation mechanisms, and safety guardrails.

Practically, this means a foreign company launching a generative AI product in China will likely go through all three: first the security assessment (the gate), then the generative AI service registration (the licence equivalent), and the algorithm filing (the ongoing registry). Understanding this stack before you start avoids the common mistake of preparing for one process only to discover another is also required.

The Role of Industry Associations and Standards Bodies

Navigating AI regulation in China is not just about interacting with government agencies. Industry associations and standards bodies play a significant role in shaping regulations, interpreting requirements, and providing guidance to companies. Understanding this ecosystem can help foreign companies stay ahead of the curve.

Key bodies include the China Artificial Intelligence Industry Development Alliance (AIIA), the China Electronics Standardization Institute (CESI), and the China Internet Network Information Center (CNNIC). These organisations develop technical standards, issue implementation guidelines, and sometimes run pilot programmes that give participating companies early insight into regulatory direction.

For foreign companies, engaging with these bodies can provide several benefits: early access to draft standards, opportunities to provide feedback, networking with industry peers and regulators, and credibility with both regulators and customers. Participation does not require a China entity in all cases — international companies can often join as observers or associate members. The level of investment you should make depends on how central China is to your business and how much regulatory risk you face.

Regulatory Arbitrage and Why It Usually Fails

Some foreign companies consider structuring their China operations in ways that try to minimise or avoid regulatory obligations — for example, by keeping all servers overseas, using a Hong Kong entity instead of a mainland China entity, or working through a local partner to avoid direct filing requirements. These approaches rarely work as well as hoped.

Chinese regulators have been increasingly willing to look through corporate structures to the substance of what is actually happening. If your service is targeting Chinese users, if it processes their data, or if it has a real connection to the China market, regulators will generally find that Chinese rules apply — regardless of where your servers are located or what entity technically operates the service.

This is not to say that structure does not matter — entity structure, data architecture, and operating model all have significant implications for exactly which rules apply and how you comply. But trying to avoid regulation entirely through structuring is usually a losing game. The better approach is to understand the rules, design your structure to comply efficiently, and build compliance into your operations from the start.

Looking at CAC enforcement actions from 2025 and the first half of 2026 reveals clear patterns about where regulators are focusing their attention. The most common violations are not grand failures — they are procedural gaps: incomplete filing descriptions, unreported algorithm modifications, and failure to display filing numbers to users.

Penalty amounts vary widely, from RMB 100,000 to several million yuan, depending on whether the violation is considered a first-time oversight or a deliberate circumvention. What is consistent across cases is that regulators expect companies to maintain their filings actively, not just submit them once and forget.

For foreign companies, the enforcement pattern has a specific implication: your compliance team needs a calendar. Algorithm change reports, annual filing reviews, and deep-synthesis label audits all have deadlines that, if missed, compound quickly. A structured compliance calendar with advance reminders is the single most cost-effective way to stay on the right side of enforcement.

Multi-Jurisdiction AI Compliance: China, EU and US Side by Side

Foreign AI companies often operate across multiple jurisdictions simultaneously, and the compliance burden multiplies quickly. China’s algorithm filing regime, the EU AI Act, and US state-level AI laws each have different philosophies, requirements and enforcement styles.

The EU AI Act follows a risk-based classification model — prohibited, high-risk, medium-risk, and limited-risk — with corresponding obligations. China’s regime is more filing-centric, with specific registration requirements for different algorithm categories. The US approach remains patchwork, with state-level laws (like California’s AI transparency rules) and sector-specific federal guidance rather than a single national framework.

For companies subject to multiple regimes, the practical strategy is to identify the areas of overlap and build a single compliance infrastructure that satisfies all of them. Data mapping, for example, is a requirement under China’s PIPL, the EU GDPR, and several US state laws. Building one robust data map that is structured to answer each jurisdiction’s questions is far more efficient than running three separate compliance projects.

Preparing for Regulator Communication: What to Expect

Foreign companies are sometimes surprised by how interactive China’s AI regulatory process is. Unlike jurisdictions where you submit a form and wait for a yes/no answer, algorithm filing and generative AI registration often involve multiple rounds of questions, clarifications, and adjustments.

The regulator’s questions typically fall into a few categories: scope clarification (is your product correctly categorised?), data sourcing (where does your training data come from?), safety mechanisms (how do you prevent harmful outputs?), and user protection (what disclosures and controls do users have?). Each round generally takes one to two weeks, and there are typically two to four rounds for a straightforward case.

The key to smooth regulator communication is preparation. Before you file, you should have clear, documented answers to the questions you are most likely to receive. This not only speeds up the process but also builds confidence with the regulator — a factor that matters more than many foreign companies realise.

FAQ

Q: How many algorithm filings are there in total?

A: As of June 2026, the CAC has announced approximately 670 algorithm filings across all categories — algorithmic recommendation, deep synthesis, generative AI services, and personalised push. This count includes only publicly announced filings; there are typically additional filings in process at any given time.

Q: How long does algorithm filing take on average?

A: The average processing time has been steadily improving. In the first half of 2026, the average was about 35 days from submission to approval. However, this varies significantly — straightforward cases can take 15-20 days, while complex generative AI cases can take 60 days or more. The total timeline from start to finish, including preparation, is usually 6-10 weeks.

Q: What percentage of filings are from foreign companies?

A: Foreign-invested companies account for approximately 4% of total filings — about 28 out of 670 as of mid-2026. While still a small share, the number has been growing steadily, from just 2 in 2022 to 28 by mid-2026. We expect this trend to continue as more international AI companies establish China operations.

Q: What are the most common reasons for filing delays?

A: The most common reasons are: incomplete or unclear algorithm descriptions, questions about training data sources and compliance, questions about safety mechanism effectiveness, user protection measure adequacy, and scope classification issues. Companies that prepare thoroughly and submit complete, well-structured filings tend to have faster approvals.

Q: Is algorithm filing a one-time requirement or ongoing?

A: Filing is an ongoing obligation. You must update your filing within ten working days of material changes to your algorithm. You also need to conduct an annual review. And as regulations evolve, you may need to adjust your filing to meet new requirements. Building compliance into your product development process is more efficient than treating filing as a one-time project.

Frequently Asked Questions↑ Back to article

How many algorithm filings are there in total?

As of June 2026, the CAC has announced approximately 670 algorithm filings across all categories — algorithmic recommendation, deep synthesis, generative AI services, and personalised push. This count includes only publicly announced filings; there are typically additional filings in process at any given time.

How long does algorithm filing take on average?

The average processing time has been steadily improving. In the first half of 2026, the average was about 35 days from submission to approval. However, this varies significantly — straightforward cases can take 15-20 days, while complex generative AI cases can take 60 days or more. The total timeline from start to finish, including preparation, is usually 6-10 weeks.

What percentage of filings are from foreign companies?

Foreign-invested companies account for approximately 4% of total filings — about 28 out of 670 as of mid-2026. While still a small share, the number has been growing steadily, from just 2 in 2022 to 28 by mid-2026. We expect this trend to continue as more international AI companies establish China operations.

What are the most common reasons for filing delays?

The most common reasons are: incomplete or unclear algorithm descriptions, questions about training data sources and compliance, questions about safety mechanism effectiveness, user protection measure adequacy, and scope classification issues. Companies that prepare thoroughly and submit complete, well-structured filings tend to have faster approvals.

Is algorithm filing a one-time requirement or ongoing?

Filing is an ongoing obligation. You must update your filing within ten working days of material changes to your algorithm. You also need to conduct an annual review. And as regulations evolve, you may need to adjust your filing to meet new requirements. Building compliance into your product development process is more efficient than treating filing as a one-time project.

Which algorithm filing categories are growing fastest in China?

As of mid-2026, deep synthesis is the fastest-growing category at 62% year-over-year growth, driven by the explosion of AI video generation tools, virtual human services, and voice cloning applications. Generative AI services are growing at 45% year-over-year as more companies launch LLM-based products and add features that trigger additional filing requirements. Algorithmic recommendation is the most mature category at 18% growth, with new filings coming primarily from new entrants rather than major platforms.

How has foreign company participation in algorithm filing evolved?

Foreign-invested company filings have grown steadily from just 2 in 2022 to approximately 28 by mid-2026, though they still represent only about 4% of total filings. The growth reflects three trends: more international AI companies establishing China entities, normalisation of the filing process making it more predictable for foreign companies, and the increasing importance of the Chinese market for AI products globally. Foreign companies typically need more preparation time than domestic ones due to entity setup, translation work, and internal alignment across global teams.

Geason Law Offices AI Industry Research Group | Reviewed by He Qianggao (Lawyer License No. 14403201310092337, 10+ years of practice)