Human Rights Watch: Interpretation of China’s Policies, Laws and Regulations (July 2026) (Issue 23)

Editor’s Note: This issue provides a focused review of the latest laws, regulations and policies concerning internet governance, legal services, digital regulation and social control. From governance of cyberbullying, information-content controls, AI interaction regulation and the “Clean Cyberspace” special rectification campaign, to rules governing lawyers’ practice and new regulations on internet information services, a series of systems have been introduced in rapid succession. On the surface, they aim to regulate online order, rectify misconduct and protect citizens’ rights and interests; in reality, taken as a whole, they show a prominent trend toward tighter controls, blurred boundaries, expanded powers and generalized discretionary authority, creating hidden pressure on such fundamental human rights as citizens’ freedom of speech, public-opinion supervision and the right to express themselves.

A review of these new rules reveals a common fatal weakness: emphasis on control over definition, and on rectification over safeguards. The vast majority of key concepts are unclear. What constitutes harmful information, oppositional speech, excessive expression, continued commentary or controversial expression? There are no quantitative standards, no clear boundaries and no exemption clauses. This crude form of legislation directly results in the unlimited expansion of enforcement authorities’ power to interpret, allowing grassroots authorities to characterize conduct arbitrarily, apply broad provisions indiscriminately and enforce selectively.

The greatest danger in the current institutional system is not that it will fail to control disorder, but that it can easily suppress the truth. Normal social criticism, reasonable public questioning, continued inquiries into public affairs and lawful public-opinion supervision may all be broadly classified as “violations involving expression,” “creating confrontation” or “disturbing order.” Vague legal provisions inherently possess the functional capacity to suppress dissent, constrain speech and punish those who speak frankly.

With eight policies layered on top of one another, the chilling effect is already plainly visible. Internet users dare not comment to avoid punishment, media outlets dare not supervise to avoid risks, and ordinary people dare not speak out to protect themselves. In the name of rectifying disorder, the actual effect is to shrink the space for speech; on the grounds of maintaining order and stability, basic civil rights are weakened. This is the most troubling aspect of the concentrated introduction of these systems.

The bottom line of the rule of law is to punish evil without harming the innocent, to remedy disorder without suppressing rights, and to regulate without blocking speech. If broad and vague rules continue to be used to expand the scope of oversight, compress lawful expression and restrict public-opinion supervision, the ultimate result will only be the contraction of public discussion, the failure of social oversight, the silencing of truth and the ossification of the public-opinion environment—seriously departing from the protection of human rights and the spirit of modern rule of law.

Contents

I. Multiple Departments Seek Public Comment for a Second Time on the Draft Revision of the Administrative Measures for Internet Information Services

II. Cyberspace Administration Launches “Clean Cyberspace” Special Campaign to Rectify Disorder in Online Entertainment Group Livestreaming

III. Half-Month Talk Publishes Analysis of Self-Media Accounts’ Disorderly “Pre-announcements” of Officials’ Downfalls

IV. Interim Measures for the Administration of Anthropomorphic Artificial Intelligence Interactive Services Officially Take Effect

V. Seven Departments Issue Opinions on Harnessing the Strength of Older People to Improve Judicial Protection for Minors

VI. Cyberspace Administration and Ministry of Emergency Management Deploy Concentrated Cleanup and Rectification of Illegal and Harmful Information Concerning Flood-Season Disasters and Accidents

VII. Authorities Publish the Draft Amendment to the Lawyers Law of the People’s Republic of China for Public Comment

VIII. Cyberspace Administration Seeks Public Comment on the Draft Anti-Cyberbullying Law

I. Multiple Departments Seek Public Comment for a Second Time on the Draft Revision of the Administrative Measures for Internet Information Services

On July 3, the Cyberspace Administration of China, together with the Ministry of Industry and Information Technology and the Ministry of Public Security, issued a notice seeking public comment for a second time on the Administrative Measures for Internet Information Services (Draft Revision for Public Comment). The comment period ends on August 2. The revision consists of six chapters and 94 articles and improves regulatory rules addressing emerging forms of business such as AI services, cyberbullying and the leakage of personal information, strengthening the foundations for protecting internet users’ personality rights, privacy rights and right to development. The draft adds several detailed provisions on human-rights protection, requiring platforms to establish early-warning mechanisms for cyberbullying and strictly control excessive collection of user information; it mandates measures to protect minors online, including the launch of dedicated youth-protection modes, to prevent cyberbullying and internet addiction. It also clarifies filing and security-assessment obligations for new services such as AI agents and algorithmic recommendations, rectifies disorders including paid online influence operations spreading rumors and false marketing, and adds a mechanism for punishing internet breaches of trust, pressing platforms to fulfill their primary responsibilities. The public may submit suggestions for revision through designated email and postal-mail channels, helping improve the rule-of-law system for cyberspace governance. (For details, see: https://www.cac.gov.cn/2026-07/03/c_1784822399677167.htm）

Analysis: The public consultation on this draft revision has positive value in terms of human-rights protection. The provisions require platforms to establish early-warning mechanisms for cyberbullying, restrain excessive collection of user information and improve protection modes for minors, thereby genuinely protecting internet users’ privacy and physical and mental health. At the same time, regulating AI and algorithmic services and rectifying rumor-spreading by paid online influence operations will purify cyberspace and build a barrier for safe internet use. The channels for the public to submit suggestions also protect citizens’ basic rights to participate in legislation and express their demands.

However, the draft has obvious shortcomings concerning freedom of expression and the balancing of rights, creating hidden risks of infringing citizens’ rights to expression and supervision. First, many prohibitive provisions are vaguely defined. Concepts such as “harmful information” and “disrupting public order” lack clear standards, leaving enormous room for discretion. To avoid punishment, platforms may adopt conservative controls. Objective and rational public criticism by internet users, and reasonable questions about government affairs, could easily be misjudged by algorithms, resulting in reduced distribution or deletion and directly shrinking the space for public-opinion supervision and freedom of speech guaranteed to citizens by the Constitution.

Second, the draft excessively increases platforms’ review obligations, forcing companies to adopt one-size-fits-all risk controls. AI identification cannot distinguish malicious attacks from objective suggestions. Ordinary people’s normal discussion of views and differentiated comments may be dealt with, causing large numbers of compliant accounts to suffer without cause and making it difficult to fully protect citizens’ right to express themselves in the digital era. At the same time, the draft focuses on establishing punishment rules while lacking detailed provisions on appeals and remedies after users’ rights are harmed. If platforms wrongly handle speech, ordinary people have narrow channels for defending their rights, and damage to personality rights and the right to express oneself is difficult to remedy.

In addition, the punishment mechanism contains a loophole of “targeting the small while letting the big off.” The draft imposes strict constraints on ordinary internet users and small and medium-sized self-media accounts, but does not impose sufficient restraints on deeper problems such as algorithmic bias and covert manipulation of public opinion by large platforms. Overall, the governance logic emphasizes control over safeguards. If implementation becomes unbalanced, it will make the public afraid to speak and erode the atmosphere of diverse online discussion. Legislation should refine the boundaries of speech and improve channels for defending rights, finding a precise point of balance between rectifying disorder and protecting citizens’ basic human rights.

II. Cyberspace Administration Launches “Clean Cyberspace” Special Campaign to Rectify Disorder in Online Entertainment Group Livestreaming

On July 3, the Cyberspace Administration of China issued the Notice on Launching the “Clean Cyberspace—Rectification of Disorder in Online Entertainment Group Livestreaming” Special Campaign, deploying a two-month, nationwide governance campaign targeting prominent violations in the group-livestreaming sector. The campaign identifies six key areas, including noncompliant account qualifications, malicious PK contests inducing large-scale tipping, gimmicks and tactics inducing consumption, vulgar and harmful content in livestreaming rooms, the illegal recruitment of minors to participate in livestreaming, and the failure of MCN agencies to fulfill their primary responsibilities. The campaign will crack down on hyping conflicts, deliberately creating sexually suggestive atmospheres, disguisedly encouraging minors to recharge and tip, and operating lowbrow account networks in bulk, while pressing platforms and talent agencies to fulfill their management obligations. Cyberspace authorities in different regions will work with platforms to improve content-review mechanisms, strictly handle violating accounts and shut down problematic livestreaming rooms, while keeping public reporting channels open. (For details, see: https://www.cac.gov.cn/2026-07/03/c_1784735652192477.htm）

Analysis: The purpose of this “Clean Cyberspace” special campaign to rectify disorder in online entertainment is to correct chronic problems in group livestreaming, including vulgar hype, induced tipping and harmful PK contests, clean up unhealthy online practices and regulate the livestreaming industry. It has positive significance for curbing disorder caused by capital and traffic manipulation and purifying the online environment for young people. However, in the algorithmic and crude implementation of grassroots platforms, the campaign has exposed major shortcomings. The most prominent problems are emphasis on punishment over identification and on control over rights protection, with the legitimate rights of many compliant ordinary creators being infringed without cause.

The greatest defect in current online governance is that rectification generally follows the principle of “better to wrongly block than to miss a violation.” Platforms rely on mechanical AI risk controls to screen content through one-size-fits-all measures, without human review, contextual differentiation or tolerance mechanisms. Entertainment interactions by legitimate teams, normal talent competitions, everyday entertainment atmospheres and compliant chat interactions can easily be crudely classified by algorithms as violations involving sexual innuendo or malicious PK contests. Large numbers of small and medium-sized accounts that follow the rules and have committed no violations have encountered inexplicable reductions in traffic, interrupted broadcasts, downgraded visibility and bans, suffering unjustified losses to their lawful creative rights.

What deserves even more reflection is that rectification often “targets the small while letting the big off.” Leading guilds and traffic teams that truly create scripted hype, organize group schemes to exploit audiences and maliciously induce large-scale tipping are difficult to target precisely because of their concealment. Ordinary independent creators without connections or schemes instead become the main targets bearing the pressure of the campaign, producing the distorted governance effect that “bad money remains while good money is punished.”

The ultimate purpose of internet regulation is to standardize the industry and protect the public, not to suppress normal creation or kill online vitality. Excessively tightening review standards, shrinking the space for lawful entertainment and disregarding ordinary creators’ legitimate rights will only make online content increasingly rigid and formulaic, stripping it of its everyday vitality. Governance should involve precise enforcement, distinctions between right and wrong and clearly defined boundaries; it should strike at genuine disorder while firmly preventing indiscriminate harm, maintaining a balance between purifying the online environment and protecting citizens’ lawful creative rights.

III. Half-Month Talk Publishes Analysis of Self-Media Accounts’ Disorderly “Pre-announcements” of Officials’ Downfalls

Issue 13 of Half-Month Talk published an in-depth investigative article directly addressing the gray industrial chain in which online self-media accounts hint at or announce officials’ impending downfall in advance, exposing prominent risks involving information leaks and traffic hype. The investigation found that many self-media accounts release signals in obscure forms, such as posting officials’ resumes and using homophones and coded language. In numerous cases, the officials involved were subsequently reported by disciplinary inspection commissions as being under investigation days after the content was posted. The article traced the sources of the information and pointed out that a small number of public officials working internally served as “insiders,” illegally leaking disciplinary-inspection and personnel leads. The relevant information was resold for profit in social-media groups, forming a complete black and gray chain. In addition to predicting officials’ downfall in advance, some accounts deliberately fabricate vulgar anecdotes and exaggerate details of alleged violations to attract traffic; even when their accounts are banned, they move to other platforms and continue spreading the content. The article warned that privately leaking information about anti-corruption investigations not only disrupts the order of investigative work and damages the seriousness of disciplinary inspection, but also easily breeds rumors and misleads public understanding. It called for strict rectification of the sources of leaks and hype-driven accounts, improvement of confidentiality mechanisms for officials’ information, and regulation of the dissemination of online anti-corruption information, in order to preserve the authority and impartiality of anti-corruption work. (For details, see: https://baijiahao.baidu.com/s?id=1870923194433688229&wfr=spider&for=pc）

Analysis: As a core Party journal sponsored by Xinhua News Agency, Half-Month Talk’s naming of the phenomenon of self-media accounts “pre-announcing officials’ downfall” has strong value in guiding public opinion and serving as a reference for discipline enforcement, and will inevitably pressure localities to regulate public-opinion management and strictly control disorder involving leaks. The phenomenon objectively exists and has complex causes. It cannot uniformly be attributed to internal leaks and profit-seeking black industries; it requires rational separation and dialectical examination. Disorder must be rectified while avoiding harm to normal public-opinion supervision.

Objectively speaking, most so-called “advance predictions” do not involve leaks of classified information. In reality, before officials are investigated, signs often appear such as vacant posts, frequent absences and a “commuting” status of remaining at home awaiting orders. Combined with public details such as homes being sealed, changes involving relatives and social rumors, neighbors and informal circles can easily detect abnormal signals. Many self-media accounts merely make comprehensive judgments based on public signs and issue cryptic hints. In essence, this is public-sector perception of public opinion, not an illegal leak. At the same time, most self-media accounts have no profit motive: advertising revenue shares for government-affairs posts are extremely low, far below the returns from ordinary livelihood and entertainment traffic. The so-called “black industry chain of hyping corruption” is clearly exaggerated, with only a very small number of accounts profiting through malicious hype.

Of course, the reason official media are targeting this type of pre-announcement is clearly that disciplinary and supervisory departments believe it causes them difficulties. First, fragmented predictions and cryptic disclosures seriously disrupt the rhythm of disciplinary investigations. Advance leaks can lead suspects to coordinate their statements or destroy evidence, interfering with the confidentiality and seriousness of anti-corruption work. Second, the dissemination of homophones, coded language and vague hints can easily generate rumors, implicate innocent officials and create unwarranted public stigma. Third, disorderly dissemination undermines the authority of official announcements, creating a distorted public-opinion environment in which “public disclosures outrun official releases” and weakening public trust in government affairs.

Moderate observation of public opinion is a supplementary force for social supervision. But governance must not be one-size-fits-all or label-driven; normal public-opinion analysis must not be equated with a black industry involving leaks. Regulators should precisely distinguish illegal internal leaks from public analysis based on publicly available information, severely punish genuine insiders who leak secrets and those who maliciously spread rumors for profit, while tolerating reasonable public-opinion supervision. While correcting unhealthy practices and regulating public opinion, they must uphold the bottom line of objective and impartial governance and protect self-media’s rights to news reporting and freedom of speech.

IV. Interim Measures for the Administration of Anthropomorphic Artificial Intelligence Interactive Services Officially Take Effect

On July 15, the Interim Measures for the Administration of Anthropomorphic Artificial Intelligence Interactive Services, issued by the Cyberspace Administration of China and four other departments, took effect, filling a regulatory gap in the field of AI emotional companionship and comprehensively protecting citizens’ personality rights and minors’ physical and mental health. The document establishes seven operational red lines, prohibiting AI-generated content involving insult and defamation, inducement to self-harm and emotional manipulation, and curbing platforms’ use of virtual personas to induce users to become addicted and spend money. The measures expressly prohibit opening intimate interactive services such as virtual partners to minors under 14, and require all relevant products to install minor-protection systems and set consumption limits and risk-warning mechanisms. The new rules also regulate user-data collection, strictly protect interaction privacy and require service providers to activate psychological-intervention channels when users display extreme emotions, balancing innovation in the AI industry with citizens’ lawful rights and interests and promoting the standardized development of anthropomorphic AI services. (For details, see https://www.gov.cn/gongbao/2026/issue_12806/202606/content_7072472.html）

Analysis: From the perspective of human-rights protection, these interim measures have positive value. The document prohibits AI-generated insults, defamation and inducement to self-harm; limits the opening of virtual intimate interactions to young minors; regulates platforms’ collection of users’ interaction privacy; protects the public’s personal dignity and right to mental health; fills the regulatory gap for emotional AI; and prevents new forms of rights violations such as emotional manipulation and deepfakes. At the same time, the measures encourage AI research and development for elder care and assistance to special groups, using technology to protect the emotional needs and right to development of people living alone and persons with disabilities.

However, the measures have obvious shortcomings in implementation and constrain citizens’ rights to expression and digital choice. First, the control provisions are broad and vague, without clearly defining the boundary between normal emotional disclosure and harmful guidance. To avoid punishment, platforms will tighten the limits of interaction. Users’ objective expression of views about social phenomena and life difficulties, as well as personalized expression of opinions, could easily be restricted, shrinking the space for online expression and exchange of ideas. Second, the blanket restriction on all anthropomorphic interactions involving younger groups ignores minors’ differing needs. Compliant AI for educational companionship and psychological counseling would also be restricted, making it difficult for minors to obtain personalized emotional assistance.

Third, the measures excessively strengthen platforms’ obligations for prior review while lacking detailed provisions on users’ remedies. If AI interaction records are sealed without justification or normal conversations are blocked without cause, users have no convenient channel for appeal, making it difficult to remedy damage to privacy and expression rights. In addition, stringent prior algorithmic review raises the research-and-development threshold for small and medium-sized enterprises, forcing diverse innovative interactive products to scale back and weakening the public’s right to independently choose digital services. Governance should refine assessment standards and improve channels for defending rights, achieving a balance between preventing AI-related rights violations and protecting citizens’ digital human rights.

V. Seven Departments Issue Opinions on Harnessing the Strength of Older People to Improve Judicial Protection for Minors

On July 17, the Supreme People’s Court, together with the Ministry of Education, the Ministry of Civil Affairs and four other departments, issued the Opinions on Giving Full Play to the Strength of Older People in Judicial Protection of Minors, improving the comprehensive human-rights protection system for minors. The opinions propose relying on retirees’ extensive life experience to carry out 10 tasks, including social investigations involving minors in cases, dispute mediation, accompaniment by suitable adults during court hearings, psychological counseling and follow-up visits after judgment. The document places particular emphasis on keeping information concerning minors involved in cases confidential throughout the process, strictly protecting their reputations and privacy, and avoiding secondary physical or psychological harm caused by case handling. Entry and exit mechanisms are established for participants, who must act voluntarily and within their abilities and provide assistance based on professional expertise in law, education, psychology and other fields. The multilateral coordination mechanism addresses shortcomings in judicial protection and builds a protective network linking the judiciary, families and society, effectively safeguarding minors’ basic rights to survival, development and protection. (For details, see: Official text: https://www.cac.gov.cn/2026-07/17/c_1785112463922461.htm）

Analysis: The opinions have positive significance for expanding channels of protection for minors and safeguarding human rights. Bringing retired legal and education professionals into case assistance can strengthen social investigations and psychological counseling, address shortages of grassroots judicial personnel, and comfort minors involved in cases through flexible means, thereby implementing minors’ rights to protection and development. The document expressly requires volunteers to sign confidentiality agreements, initially reducing the risk of case information leaks institutionally and reducing secondary damage to minors’ reputations.

However, the implementation of the new policy contains multiple potential rights risks, and its shortcomings require urgent attention. First, older volunteers are not formal judicial personnel and lack systematic professional training in protecting minors’ rights and interests. Some retirees may rely on outdated assumptions when providing family-education guidance and conducting social visits, easily making subjective judgments about minors while ignoring young people’s independent personalities and rights to autonomous expression, thereby causing secondary psychological trauma. Second, the admission-review mechanism contains loopholes. It establishes only basic checks for prior criminal records and lacks regular psychological assessments and dynamic supervision. Volunteers have access throughout the process to family information and case details involving minors. If their awareness of confidentiality is weak, the privacy and identity information of minors involved in cases could easily leak, completely violating the basic principle of protecting minors’ privacy.

Furthermore, the opinions do not clearly divide volunteers’ powers and responsibilities, and some visits and mediation work may not be accompanied throughout by public officials. When dealing with minors who have been harmed, improper questioning by nonprofessionals may deepen their psychological scars; if volunteers make biased statements, they may indirectly harm minors’ rights to reputation. At the same time, the document lacks a channel for objections by victims. If guardians are dissatisfied with the way volunteers work, it may be difficult for them to request a replacement or termination of services. Expanding protection channels cannot come at the cost of weakening rights safeguards. Subsequent measures must refine supervision, objection and accountability mechanisms and uphold the core human-rights protections for minors.

VI. Cyberspace Administration and Ministry of Emergency Management Deploy Concentrated Cleanup and Rectification of Illegal and Harmful Information Concerning Flood-Season Disasters and Accidents

On July 23, the Cyberspace Administration of China and the Ministry of Emergency Management jointly issued a notice deploying a nationwide campaign to clean up and rectify illegal and harmful information concerning disasters and accidents during the flood season. The campaign aims to block the spread of false information concerning disasters, regulate the release of online information during the flood season, and prevent false messages from misleading the public and disrupting rescue and relief work. It focuses on five categories of problems: maliciously editing and splicing videos of disasters elsewhere and presenting them as local emergencies; recycling old disaster footage and falsely claiming it depicts current events; fabricating or exaggerating disaster data; inventing emergencies such as dam collapses and leaks and falsely issuing notices in the name of authorities; and using AI to fabricate disaster materials or carry out online fraud under the guise of disasters. The notice requires cyberspace authorities in all localities to conduct comprehensive investigations of all types of harmful information and urge online platforms to fulfill their primary responsibilities. Emergency-management authorities will strengthen cross-departmental coordination. The two departments urged internet users to consciously resist and not believe online rumors concerning disasters and to work together to create a clean cyberspace. (For details, see: https://www.cac.gov.cn/2026-07/23/c_1786465620398126.htm）

Analysis: The two departments jointly deployed a special campaign to rectify harmful information concerning disasters during the flood season. From the perspective of top-level design, the policy’s purpose is beyond doubt: to combat malicious online disorder such as AI-generated fabrications, spliced disaster footage, fabricated emergencies and incitement to panic; safeguard flood-control and disaster-relief order; and protect public safety and social stability. It has legitimate governance value. Yet in the reality of grassroots governance, the campaign may easily become distorted, with its original purpose seriously diverging from implementation, exposing deep-rooted contradictions in online governance and the public disclosure of disaster information.

The central authorities have consistently required disaster information to be truthful, accurately reported and transparent, and have strictly prohibited concealment or underreporting. However, under the pressure of grassroots assessments and official career advancement, some local officials have a strong subjective motivation to understate disasters, reduce reported casualties and conceal losses in order to avoid accountability and protect their achievements and promotion prospects. This creates a real paradox: official public data may not necessarily be accurate, while firsthand observations from the public may be closer to the truth.

Against this background, one-size-fits-all public-opinion rectification can easily produce a double injustice. First, ordinary self-media accounts, limited by their information channels, may publish data that differs slightly from official announcements without any malicious intent to spread rumors, yet are routinely classified as disseminating false information and have their accounts dealt with. Official media, by contrast, may occasionally publish inaccurate or erroneous reports and resolve the matter merely through corrections or withdrawals, with almost no accountability, creating a double standard in which “official errors can be forgiven, while public errors must be punished.”

Second, the greatest danger of this campaign is that it could become a tool for local authorities to conceal disasters and suppress public-opinion supervision. Once a locality has concealed or underreported a disaster, self-media accounts that disclose the true situation or expose conditions at the scene may be cleaned up or banned for being “inconsistent with the official account.” Suppressing the truth while protecting falsehoods not only seriously harms press freedom and the public’s right to know, but also makes it difficult for higher authorities to identify grassroots governance weaknesses through public opinion, completely departing from the original goal of seeking truth from facts.

More troublingly, the campaign could easily become a “compliance tool” for some grassroots cyberspace authorities, giving them an opportunity to settle scores with or restrict self-media accounts that have long spoken independently and boldly and specialized in supervision, thereby creating a chilling effect on online expression.

Truly sound governance should crack down on malicious rumor-spreading for profit, tolerate public voices that contain information discrepancies, and strictly investigate local concealment or underreporting of disasters. Only by restraining the hands of authorities that conceal information and opening the mouths of the public to supervision can a positive cycle of clear public opinion and transparent disaster relief truly be achieved.

VII. Authorities Publish the Draft Amendment to the Lawyers Law of the People’s Republic of China for Public Comment

On June 26, the National People’s Congress website published the Law of the People’s Republic of China on Lawyers (Draft Amendment) and formally sought public comment nationwide. The deadline for comments is July 25. The draft was reviewed by the 23rd meeting of the Standing Committee of the 14th National People’s Congress. The public may submit comments online through the National People’s Congress website (www.npc.gov.cn), or mail written suggestions to the Legislative Affairs Commission of the Standing Committee of the National People’s Congress. This amendment places equal emphasis on protecting lawyers’ professional rights and regulating professional conduct, and many provisions concern judicial human-rights protection. The draft makes clear that public-security, procuratorial and judicial organs and other authorities should improve their systems and protect lawyers’ professional rights, including meeting clients and reviewing case files, in accordance with the law. It promotes full coverage of lawyer defense in criminal cases through legislation, better protecting defendants’ right to defense. It also improves provisions concerning foreign-related and public-interest legal services and refines rules on entry into the legal profession and industry supervision. Industry insiders believe that this revision will further strengthen the foundations of the lawyer system and continuously improve mechanisms for fair justice and human-rights protection. (For details, see: https://news.ynet.com/2026/06/28/4021921t70.html） (Official public-comment portal:

https://www.npc.gov.cn/npc/c30834/202606/26/content_10623478.html）

Analysis: Examined through the underlying logic of modern judicial human rights, this revision of the Lawyers Law displays an extremely clear one-sided legislative character: constraints and punishments are comprehensively upgraded, while rights safeguards seriously lag behind. Although it appears to improve industry governance and strengthen judicial protections, it has failed to cure the core and most fatal chronic human-rights problem in China’s criminal-defense field—the unclear boundaries of professional safety for independent defense, leaving outspoken lawyers highly vulnerable to collateral harm and retaliation.

Compared with the old law, the progress in this draft remains limited to superficial universal benefits. By institutionalizing full coverage of criminal defense and detailing the obligations of public authorities to provide protection, it somewhat fills gaps in the judicial remedies available to disadvantaged parties, allowing ordinary people to “afford a lawyer and have an opportunity to defend themselves.” Increasing punishment for bad actors in the profession who engage in malicious hype, illegal practice or false representation is also necessary to purify the rule-of-law environment; strict regulation of lawful and compliant conduct is beyond reproach.

But the three core provisions that truly touch the heart of judicial human rights—professional immunity, tolerance for conduct in the performance of duties and immunity for speech—are completely avoided in this amendment, exposing the conservatism and shortcomings of the legislation. For a long time, China’s judicial practice has contained a cruel and seemingly insoluble paradox: public power enjoys broad tolerance for error, while private lawyers have zero tolerance in performing their duties. Defects in evidence-gathering, factual deviations and procedural omissions by official case handlers are mostly concluded through internal rectification without accountability. But once a lawyer adheres to legal principles, dares to challenge flaws in an investigation, questions gaps in the chain of evidence or insists on a defense of innocence, the lawyer may easily be labeled as “obstructing justice” or “inciting interference,” facing interviews, revocation of a license or criminal prosecution—even when acting entirely on the basis of facts and law.

Over the decades, many practicing lawyers with ideals, integrity and faith in the rule of law have not violated laws or disciplines; they simply refused to blindly follow predetermined conclusions, dared to speak independently and insisted on examining case details, ultimately ending up imprisoned. This is not an isolated enforcement error, but a systemic injustice caused by vague legal boundaries.

Most regrettably, this draft amendment continues the old defect: it only specifies what lawyers may not do, without clearly establishing the bottom line of absolute immunity for lawyers performing their duties lawfully. It does not refine rules on absolute immunity for statements made in court, establish a rigid boundary distinguishing “lawful independent defense” from “unlawful interference with justice,” or create a tolerance mechanism for rigorous defense and professional challenges.

This means that major loopholes will remain after the amendment: the malicious illegal conduct of bad lawyers may not be punished precisely, while the independent and outspoken conduct of good lawyers may still lead to criminal charges at any time. Excessive grassroots judicial discretion can easily equate the adversarial nature of professional defense directly with opposition to public power, indirectly suppressing lawyers’ independence in defense.

The essence of judicial human rights has never been to “restrict lawyers’ speech,” but to protect lawful professional conduct, tolerate professional adversarial practice and prevent retaliation by those in power. A mature Lawyers Law should be a protective shield for those who speak boldly, not a restraint on those performing their duties. This revision emphasizes “control” over “protection” and punishment over tolerance for error. It has failed to dispel criminal-defense lawyers’ fear in practicing their profession and has failed to fundamentally build the final safeguard for citizens’ right to defense, a basic human right. Unless the shortcomings concerning immunity are addressed and the boundaries of professional duties clarified, so-called judicial protection will ultimately remain merely formal.

VIII. Cyberspace Administration Seeks Public Comment on the Draft Anti-Cyberbullying Law

On July 29, the Cyberspace Administration of China issued a notice seeking public comment nationwide on the Law of the People’s Republic of China Against Cyberbullying (Draft for Public Comment). The deadline for comments is August 28. The full text consists of seven chapters and 60 articles. For the first time, it defines cyberbullying through special legislation, establishes a complete governance chain of prior warning, intervention during incidents and accountability afterward, strengthens internet platforms’ primary responsibilities, improves mechanisms protecting victims’ rights, and draws clear legal boundaries for conduct including rumor-spreading and defamation, doxxing and coordinated online attacks. The public may submit comments by email or letter before the deadline. (For details, see: https://www.cac.gov.cn/2026-07/29/c_1787072711938509.htm）

Analysis: This draft for public comment directly addresses the rampant cyberbullying of doxxing, malicious attacks, rumor-spreading and humiliation. Special legislation to protect citizens’ personal dignity and respond to public expectations is undoubtedly a good measure for safeguarding ordinary people’s basic human rights. But the provisions still contain areas of ambiguity, placing a difficult question before everyone: How can perpetrators of cyberbullying be precisely targeted while avoiding harm to lawful public expression and preventing the contraction of the space for speech, so that internet users do not become too frightened to speak when problems arise?

The draft defines “centrally and continuously publishing information that incites confrontation and infringes upon others’ rights and interests” as cyberbullying, but its core concepts lack clear quantitative standards. What does “centrally” mean? How many comments, or what scale of dissemination, qualifies as “continuous”? How can “inciting confrontation” be distinguished from objectively exposing livelihood problems and deliberately creating social division? Vague definitions can easily create space for flexible enforcement. Once the standards are applied unevenly, rational questions and sustained public-opinion supervision concerning public affairs and public officials could be simply brought within the scope of regulation.

Freedom of speech is an important basic right of citizens, and public-opinion supervision is an indispensable channel for society’s self-purification. Sharp criticism is not cyberbullying, and persistent questioning is not malicious attack. In reality, many public voices are grounded in facts, focus on shortcomings in grassroots governance and constitute lawful and reasonable expression of public opinion. If legal boundaries are vague, the provisions risk being used selectively and becoming tools for restricting criticism and suppressing those who speak frankly. When ordinary people fear that their rational questions may be deemed to have “crossed the line,” the ultimate result will be that the public dares not speak and channels for social supervision shrink.

The goal of governing cyberbullying should be to eliminate malicious harm such as insults, threats and doxxing, not to eliminate different voices. Good laws lie in precision: they must neither allow cyberbullying to trample human rights at will nor, for the sake of administrative convenience, erect overly broad barriers to expression. Legislators need to further refine the standards for distinguishing among types of conduct and add clear exceptions: public-opinion supervision carried out in accordance with the law, and fact-based reporting and criticism, should not be subject to restrictions under this law.

The vitality of law lies in balance. It is hoped that during this round of public consultation, all sides will fully debate and refine the provisions, drawing a clear dividing line: allowing wanton cyberbullying to be severely punished while ensuring that reasoned and evidence-based public expression has a safe space. Only by taking both the protection of personal rights and freedom of speech into account can this law truly achieve good law and good governance.