Zhuo Sheng Original | On the Application Conflict and Improvement of Articles 5 and 24 of the Quality and Effectiveness Standards for Review and Arrest in 2026
Author | Mak Cheuk-lam, Zhuosheng Criminal Law Team
Abstract
The “Standards for the Quality and Efficiency of Arrest Review by People’s Procuratorates (Trial)” issued in 2026 is an important normative document through which procuratorial organs implement the criminal justice policy of minimizing arrests, prosecutions, and detention, and standardize the criteria for handling arrest review cases. The new regulations clarify the two distinct review standards for arrest and prosecution, refine the criteria for assessing social dangerousness, and expand the scope of defense counsel’s participation during the arrest-review stage. These measures hold significant practical value in standardizing the application of detention and reducing the rate of unjustified arrests.
However, in practical implementation, the core guidance provided in Article 5—regarding the differentiation between arrest and prosecution standards—clearly conflicts with the case review rules for improperly approved arrest cases set out in Article 24: Article 5 permits a procedural logic whereby an arrest may be authorized if the evidence meets the threshold, even if subsequent evidence proves insufficient to support a prosecution; yet Article 24 broadly includes post-arrest non-prosecutions within the scope of adverse quality assessments, thereby creating widespread concerns among prosecuting officers about potential accountability. As a result, prosecutors have, in effect, reverted to the traditional practice of using prosecution standards in lieu of arrest standards, undermining the original intent of the new regulations to establish distinct evidentiary thresholds for the two stages.
Drawing on practical experience in case handling, the author compares the statutory requirements for arrest and prosecution, distilling a core litigation logic of eighteen characters: “arrest based on evidence first, followed by non-prosecution; prosecute after initial non-arrest based on circumstances.” Furthermore, by examining a real‑world case involving the issuance of fictitious invoices, the author illustrates how conflicts among statutory provisions can harm the rights and interests of the parties involved. In light of this, the author proposes a feasible amendment to refine and improve the review criteria under Article 24, while also outlining a practical, detailed defense strategy for pre‑trial detention that attorneys can implement under the new regulatory framework, aiming to strike a balance between ensuring the quality of case handling and preserving the independent discretion afforded to coercive measures.
1. Standardized interpretation of the original text:
Article 5: Defining the Boundaries for the Independent Application of Arrest Standards
(I) Article 5 in its entirety, along with the two core principles it embodies.
1. Original text of the provision
Article 5:Consistently and accurately apply the legal standards for arrest. Strengthen the review of evidence in cases involving arrest reviews, accurately assess the factual and evidentiary requirements for arrest, the sentencing requirements, and the criteria for social dangerousness, and avoid conflating the standards for arrest with those for prosecution. It is neither permissible to use the standard for reviewing and prosecuting as the standard for arrest, thereby deciding not to arrest a criminal suspect who meets the criteria for arrest but has not yet reached the threshold for initiating public prosecution, nor is it permissible to lower the arrest standard and arrest a criminal suspect who does not meet the criteria for arrest.
2. Interpretation of the two-tier rigid constraints in the provision
This provision serves as the overarching principle governing the standards for arrest and prosecution throughout the entire quality and efficiency framework. It directly addresses two longstanding issues in judicial practice—namely, the artificial elevation of arrest thresholds and the blanket practice of arresting anyone who merely constitutes a crime—and sets forth two equally important requirements: first, it prohibits the arbitrary raising of the threshold for reviewing arrest applications. The stringent standard of “sufficient and conclusive evidence” applicable to the review and prosecution stage should not be mechanically applied to assess the evidence at the arrest stage. As long as a case meets the three statutory requirements for arrest, even if the existing evidence does not yet satisfy the standards for public prosecution, an arrest decision should not be withheld on that ground. The primary function of the arrest procedure is to ensure the smooth conduct of investigations and proceedings, rather than to prematurely complete a conviction review.
Second, it is prohibited to unilaterally and excessively relax the conditions for issuing arrest warrants. We must abandon the outdated, overly broad approach of “arresting anyone who commits a crime.” For criminal suspects who pose no public safety risk, have committed minor offenses, and meet statutory conditions for leniency, even if their actions constitute a crime, a decision not to arrest should be made in accordance with the law, and the scope of pre-trial detention must be strictly limited.
(II) A comparison of the three core differences between arrest standards and prosecution standards.
Article 7 of the new regulations systematically sets forth the three statutory conditions for arrest, while Article 7 of the 2026 Standards for the Quality and Efficiency of Prosecutorial Review specifies the substantive criteria for prosecution. These two provisions exhibit a hierarchical differentiation in terms of evidentiary requirements, sentencing discretion, and the focus of review, which also constitutes the legal basis for the author’s summarized eighteen-character case-handling logic. The following discussion will present a layered comparison.
1. There is a significant disparity in the hierarchical standards for evidentiary proof.
Arrest only requires “evidence demonstrating the existence of a criminal fact”; such evidence need merely provide preliminary corroboration and be sufficient to identify the suspect, without requiring the elimination of all reasonable doubt. In contrast, the review for prosecution requires that “the criminal facts are clear, and the evidence is true and sufficient, with all reasonable doubt eliminated,” representing a significantly higher standard of proof than that for arrest.
Due to the gradient difference between the evidentiary standards for arrest and prosecution, there naturally exists a category of intermediate cases: the existing evidence meets only the threshold for arrest, but even after the investigative authorities have exhausted all supplementary investigation measures, key evidence still contains defects and fails to satisfy the standard for public prosecution. In accordance with the explicit provisions of Article 5, in such cases, arrest should be carried out first to ensure the smooth conduct of the investigation; if the evidence cannot be supplemented during the subsequent review and prosecution stage, a decision of non-prosecution shall be made in accordance with the law. This constitutes the normative basis for the practice of “arresting first on the basis of evidence, followed by non-prosecution.”
2. The scope of application for sentencing discretion varies in breadth.
The threshold for applying arrest as a coercive measure is merely “a potential sentence of fixed-term imprisonment or above,” while also establishing a tiered discretion framework: suspects who may face a sentence of ten years or more, have a prior record of intentional crimes, or have violated the regulations governing release on bail are prioritized for arrest. In other words, for cases involving minor offenses punishable by less than ten years’ imprisonment, with no prior criminal record and where there is no likelihood of violating bail conditions, case-handling authorities have ample discretion to refrain from arresting the suspect.
In contrast, in judicial practice, the standard scope of non-prosecution is largely confined to cases where the potential sentence is three years’ imprisonment or less, the circumstances are minor, and compensation has been made with the victim’s forgiveness. Thus, the scope of application for non-arrest is much broader than that for non-prosecution. A large number of minor criminal cases can be subject to a decision of non-arrest, yet, considering all the circumstances of the case, they still meet the conditions for public prosecution, which aligns with the practical logic of “non-arrest first, followed by prosecution.”
3. The core focus of the review and evaluation is completely distinct.
Detention review is a procedural coercive measure review, with its focus on assessing future risks—specifically, whether the suspect poses a social danger of fleeing, colluding with others to fabricate evidence, or committing further crimes if released on bail. In contrast, prosecution review is an examination of the substantive elements for conviction and sentencing, focusing on past criminal facts and comprehensively evaluating such substantive factors as the social harm caused by the criminal act, the subjective malice of the perpetrator, and the methods and consequences of the crime.
In simple terms, arrest focuses on whether further problems are likely to arise, while prosecution assesses whether past criminal conduct should be held accountable. The evaluation criteria for the two are distinct and do not overlap, further demonstrating that the two sets of standards should not be applied interchangeably.
(III) The eighteen-character core logic: in terms of evidence, arrest first then no prosecution; in terms of circumstances, no arrest first then prosecution.
In light of the differences among the aforementioned three-tier standards, and drawing on my extensive experience in criminal defense practice, I have distilled an eighteen-character summary that encapsulates the core essence for interpreting the spirit of Article 5 and distinguishing between the criteria for arrest and prosecution. This succinct formulation comprehensively outlines the two lawful procedural pathways envisioned by the new regulations:First, in terms of evidence: arrest first, then decline to prosecute.The evidence in the case only meets the low threshold for arrest but has not yet satisfied the higher standard required for prosecution. In accordance with Article 5, a decision not to arrest cannot be made directly; an arrest shall be ordered in accordance with the law. If, after two rounds of supplementary investigation and the exhaustion of all investigative measures, the evidence still cannot be substantiated and fails to meet the prosecution standard, a decision of non-prosecution may be rendered in accordance with the law.
Secondly, in terms of the sequence of events, there is an initial decision not to arrest followed by a subsequent prosecution:The suspect’s offense is minor, has no prior criminal record, has voluntarily surrendered and returned the ill-gotten gains, and poses no objective public safety risk; thus, the statutory conditions for non-arrest are met, and a decision not to arrest should be made. Even if the suspect is not detained, considering the totality of the criminal facts and circumstances in this case, there remains a justifiable need to prosecute, and a public prosecution shall be instituted in accordance with the law during the subsequent review and prosecution stage.
(IV) The contrasting practical implementation of the two logics
1. “In terms of the case details, arrest is withheld initially, followed by prosecution.”
Implementation has been smooth and aligns with the policy of minimizing arrests. For minor offenses involving prostitution, gambling, drugs, and civil economic disputes—particularly in cases where the suspect voluntarily offers compensation, obtains the victim’s forgiveness, and is a first-time or occasional offender—most frontline prosecutors are now able to apply the spirit of the new regulations by issuing decisions not to arrest on the grounds of no social danger, followed by standard prosecution based on the facts of the crime and the necessity of prosecution.
This approach strikes a balance between lenient detention and crime-fighting, and does not encounter any significant obstacles in case handling.
2. “Arrest first, then decline to prosecute based on the evidence”
It is almost at a standstill, and its implementation has been severely hindered. In practice, very few prosecutors are willing to conduct cases under Article 5 by first arresting the suspect and then deciding not to prosecute. Even if the case evidence only marginally meets the arrest standard, and there is a high likelihood that it will not subsequently be sufficient to meet the prosecution standard, the handling prosecutor will deliberately raise the evidentiary threshold for arrest and simply decline to authorize arrest on the grounds of insufficient evidence. If, due to the nature of the case, an arrest decision must be made, the prosecutor will then devote all efforts during the review‑for‑prosecution stage to securing and consolidating the evidence, proceeding with the prosecution regardless, and rarely issuing a decision not to prosecute.
The direct institutional obstacle causing this phenomenon is the negative case review rule stipulated in Article 24 of the Quality and Efficiency Standards.
2. The root cause of statutory conflicts:
Article 24 of the review and appraisal rules undermines the original legislative intent of Article 5.
(I) Article 24: Full original text and institutional positioning
1. Original text of the provision
Article 24:A case is deemed an unqualified arrest if any of the following circumstances exist: (I) approval or a decision to arrest a person who has no criminal facts or falls under the circumstances specified in Article 16 of the Criminal Procedure Law where criminal responsibility is not pursued; (II) after arrest, the case is withdrawn, prosecution is not initiated, or the court renders a verdict of not guilty, and a review of case quality confirms that the arrest was seriously erroneous; (III) approval of arrest for a criminal suspect who is unsuitable for detention and for whom arrest is unnecessary; (IV) approval of arrest by this court despite lacking jurisdiction.
2. The original intent behind the provision and its inherent defects
The purpose of this provision is to establish a mechanism for negative evaluation of arrest cases, thereby regulating unlawful arrest practices such as arrests made without factual basis, arrests that are not necessary for detention, and arrests exceeding jurisdictional authority. This mechanism aims to compel prosecutors to exercise their power of approving arrests with due diligence, and to prevent wrongful arrests and improper detention. However, the legislative drafting contains two critical loopholes that directly create a logical conflict with Article 5: first, the non-prosecution circumstances are not hierarchically differentiated, adopting a “result-based liability” approach. Article
merely states in general terms that “cases in which a person is arrested but not prosecuted, and the arrest is subsequently determined through review to have been a serious error, shall be deemed non‑compliant,” without distinguishing the underlying causes of the non‑prosecution. It encompasses both cases of wrongful arrest—where the factual findings at the arrest stage are entirely erroneous and unsupported by any evidence—and cases of lawful arrest followed by non‑prosecution due to the failure to supplement the evidence during subsequent investigation. Both types of cases are uniformly included within the scope of adverse review.
Second, the criterion of “an arrest that is demonstrably and seriously erroneous” lacks detailed, quantifiable standards for determination, leaving too much room for discretionary judgment. The new regulations have not been accompanied by interpretive documents, nor do they specify which circumstances constitute “serious errors.” As a result, case-handling personnel are unable to clearly distinguish between “arrest followed by non-prosecution in accordance with the law” and “illegal or erroneous arrest,” and are thus prone to developing, on a subjective level, a sense of anxiety—namely, the fear that any instance of arrest followed by non-prosecution will be deemed evidence of substandard case handling.
(II) Under the integrated prosecution and arrest model, the conflict between two provisions gives rise to practical difficulties.
Currently, procuratorial organs nationwide are comprehensively implementing the “arrest-and-prosecution integration” case-handling mechanism, under which the same prosecutor or case-handling team is responsible throughout the entire process for reviewing arrests and reviewing prosecutions in a single case. The performance evaluation, quality review, and accountability of case-handling personnel are all tied to the full lifecycle of that single case. The outcome-oriented review rules set forth in Article 24 directly reshape prosecutors’ case-handling mindset: from the perspective ofPerforming one's duties risk, once an arrest decision is made, if a decision not to prosecute is subsequently rendered, the case will be subject to a quality review procedure. Even if it is ultimately not determined to constitute a “serious error,” the handling prosecutor must still prepare extensive explanatory reports and undergo retrospective reviews and verifications, thereby incurring a substantial additional workload. Moreover, if the review determines that the arrest decision involved a serious error, it will directly affect the individual’s annual performance appraisal, eligibility for commendation and promotion, and may even trigger an investigation into judicial liability.
Driven by their professional instinct to maximize benefits and minimize risks, the vast majority of prosecutors will proactively avoid the risk of declining to prosecute after an arrest and instead raise the arrest threshold directly to the prosecution standard: if the evidence fails to meet the standard for public prosecution, they will refrain from arresting; once an arrest is approved, regardless of how the circumstances subsequently evolve, they will invariably proceed with the public prosecution.
This conservative approach to case handling effectively renders Article 5—prohibiting the conflation of arrest and prosecution standards—a mere formality, thereby undermining the practical significance of the new regulations’ institutional design that distinguishes between two sets of evidentiary standards. Even if prosecutors inwardly recognize the normative spirit of Article 5, they will not proactively assume the risks associated with review and accountability; in the “eighteen-character logic,” the pathway of “arresting first on the basis of evidence, then declining to prosecute” is completely blocked.
(III) Summary of the underlying logic of the conflict
Article 5 establishes a positive case-handling guideline for the “independence of arrest standards,” granting prosecutors lawful discretionary authority to apply arrest based on low evidentiary thresholds and recognizing the objective reality that evidence tends to be progressively refined during the course of litigation. Article 24, by adopting a single‑outcome‑oriented quality‑assessment logic and failing to include tiered exculpatory provisions, categorizes all post‑arrest decisions not to prosecute as negative outcomes. These two provisions create a direct conflict in their evidentiary discretion rules, thereby constituting an internal inconsistency within the regulatory framework.
3. Supporting practical cases:
Cases of issuing false invoices vividly demonstrate the real-world harm caused by conflicts among legal provisions.
(I) The complete basic facts of the case
The law firm where the author works is handling a criminal case involving the illegal issuance of invoices, with the face value of the invoices in question amounting to approximately RMB 1.5 million, which falls within the statutory sentencing range of up to two years’ imprisonment; the suspect voluntarily surrendered and thus qualifies for the mitigating circumstance of self-surrender. During the review‑and‑arrest stage, the suspect truthfully confessed to all the acts of issuing false invoices but failed to proactively disclose the details regarding the illicit gains of over RMB 20,000. Based on this, the prosecuting officer determined that the suspect had not fully and truthfully confessed, and therefore did not recognize the confession as voluntary surrender. The officer may have also considered that there was a risk of interference with witnesses or destruction of evidence, and thus issued a decision approving the arrest.
After the case was transferred for review and prosecution, the suspect fully confessed to the facts of obtaining illicit gains, voluntarily signed a statement of guilt and acceptance of punishment, and turned over all illegal proceeds in full. There are no aggravating circumstances; moreover, the suspect qualifies for multiple statutory and discretionary mitigating factors, including voluntary surrender, being a first-time offender, and full restitution of the ill-gotten gains. However, the prosecuting officer, having already issued an arrest decision in the earlier stage, in order to avoid the risk of failing the case review under Article 24, did not issue a decision not to prosecute and instead still brought the public prosecution before the court.
During the first-instance trial, the defense counsel submitted a comprehensive legal opinion, and the court ultimately accepted the defense’s argument regarding voluntary surrender, sentencing the defendant to six months’ imprisonment.
(II) Analyze the procedural deficiencies in the handling of this case in light of the new regulations.
1. In accordance with Article 28, this case should have been decided not to be prosecuted in accordance with the law.
Article 28 of the new regulations clearly specifies seven mandatory circumstances under which arrest should not be authorized. As long as these conditions are met, the procuratorial organs have no discretion and must refrain from arresting; these include cases involving first-time offenders with relatively minor subjective malice, those who voluntarily surrender after committing the crime, and those who actively return stolen goods or make restitution. The suspect in this case fully meets the aforementioned criteria and constitutes a typical low‑severity economic crime; there is no real risk that they will continue to issue false invoices, collude with co‑defendants, or flee. Moreover, the new regulations require public security organs to submit objective evidence demonstrating the existence of social danger; they may not rely solely on subjective speculation to conclude that a suspect poses a risk of collusion. In this case, the determination of social danger lacks objective evidentiary support, and the arrest decision itself is open to question.
2. Conflicts among statutory provisions directly result in irreparable harm to the parties’ rights and interests.
The suspect’s multiple mitigating circumstances fully satisfy the conditions for non-prosecution; thus, a criminal record could have been avoided through a decision not to prosecute. However, because the arrest decision was made first, and the case-handling personnel are bound by Article 24 of the evaluation rules, they dare not make a decision not to prosecute. Ultimately, the individual was convicted and sentenced, leaving a permanent criminal record that has long-term adverse effects on their employment prospects, creditworthiness, their children’s political background checks, and business activities, resulting in an excessive litigation burden.
This case clearly demonstrates that the normative conflict between Article 5 and Article 24 directly narrows the scope for lenient treatment of defendants charged with minor offenses, thereby contravening the core human rights‑oriented objective of the policy of fewer arrests, cautious prosecution, and prudent detention.
(III) Extended Reflection: The subjective discretion in the review of social dangerousness has been tightened by the new regulations.
In such minor, economically motivated offenses, the determination of whether a suspect poses a public safety risk inherently involves a degree of subjective judgment, and in the past, case-handling personnel have often relied solely on subjective assumptions to make detention decisions. However, following the issuance of the 2026 standards on quality and effectiveness, the burden of proof regarding social dangerousness is explicitly placed on the investigative authorities. Where there is no objective written evidence to support the risks of flight, recidivism, or interference with testimony, the requisite level of social dangerousness for arrest may not be deemed to exist. This significantly narrows the margin of discretion available to prosecutors and provides a solid normative basis for defense counsel in mounting defenses against arrest.
4. Path to improving the regulatory system:
Refine Article 24 to ensure seamless alignment with Article 5.
In response to the internal logical conflict between the two provisions, the core solution is to revise and refine Article 24 of the review rules by introducing tiered criteria for determination and statutory exculpatory circumstances, thereby alleviating prosecutors’ concerns about performing their duties and ensuring that the legislative intent underlying Article 5—distinguishing between arrest and prosecution standards—is effectively implemented. Specifically, three feasible recommendations for improvement are put forward:
(I) Categorize and define the grounds for non-prosecution, and introduce an exemption clause for compliance-based non-prosecution after arrest.
Amend the wording of Article 24, paragraph 2, to adopt a differentiated evaluation approach for cases where no prosecution is initiated after arrest, categorizing them into two types: First, cases falling under exculpatory circumstances (not to be deemed as improper arrests): where, at the arrest stage, the evidence was sufficient and all three statutory conditions for arrest were fully met, but the investigative authorities subsequently failed to supplement key evidence and thus could not meet the prosecution standards, resulting in a decision not to prosecute. Such cases are considered lawful law-enforcement actions pursuant to Article 5 and shall be excluded from the scope of review for improper arrests, thereby exempting the case-handling personnel from associated accountability risks.
Category II: Negative review circumstances (classified as a serious error): If an arrest lacks any factual basis or evidentiary support, and the arresting authority knowingly proceeds with the arrest despite knowing that there is no public safety risk and no criminal facts, and the case is subsequently dismissed without prosecution, then the arrest shall be deemed non‑compliant, and the quality accountability procedure shall be initiated. By refining the provisions, a normative basis is directly provided for prosecutors to implement the eighteen-character logic of “arrest first, then decline to prosecute based on evidence,” thereby eliminating concerns about liability in the course of performing their duties.
(II) Issue accompanying explanatory documents to refine the checklist for determining "a serious error in the arrest."
The Supreme People’s Procuratorate has concurrently issued the “Guidance on the Application of Standards for the Quality and Effectiveness of Arrest Reviews,” explicitly listing all circumstances constituting a “serious error in arrest,” such as: approving an arrest in the absence of any criminal facts; approving an arrest despite the lack of objective evidence to substantiate social dangerousness; continuing detention when it is known that the offense is minor and statutory grounds for non-arrest apply; or relying solely on illegally obtained, coerced evidence as the basis for arrest.
By establishing a clear and concrete negative list, the boundary between “lawful discretion” and “illegal wrongful arrest” is delineated, thereby preventing case-handlers from relying solely on subjective assumptions out of fear of performance evaluations and accountability. This ensures that prosecutors have clear regulatory guidance when applying Article 5.
(III) Reform the single outcome-oriented review model and establish a comprehensive, full-process evaluation mechanism.
We should move away from the current outcome-based evaluation approach that retroactively determines whether an arrest was erroneous solely based on post-arrest decisions such as non-prosecution or acquittal, and instead establish a tiered, full-process review system. Case quality assessments should not be limited to the final litigation outcome; rather, they should comprehensively examine all materials generated during the arrest phase, including interrogation transcripts, evidence verifying social dangerousness, outlines for continued investigation, and documents responding to counsel’s opinions.
As long as, during the arrest phase, a thorough substantive review is conducted in strict accordance with Articles 5 and 7—covering evidence, sentencing considerations, and the degree of social danger—and the rationale for the arrest is fully documented along with clear guidance on the direction of subsequent investigative evidence‑gathering, even if a decision is ultimately made not to prosecute, the arrest will not be deemed to have involved a serious error. This approach ensures that procedural compliance is assessed independently of the substantive outcome of the litigation.
5. Practical Approaches to Refined Criminal Defense Under the New Regulations
Following the implementation of the 2026 standards for the quality and effectiveness of arrest reviews, criminal defense during the arrest review phase has completely moved away from the traditional model that relied on experience, generic templates, and informal networking, and has officially entered a new era ofRefined defense—characterized by rule-based practice, evidence‑based arguments, and substantive adversarial proceedings. In light of all the provisions of the new regulations, as well as the practical implications arising from the conflict between Article 5 and Article 24, lawyers can develop a layered, systematic defense strategy and fully leverage the defense rights conferred by the new regulations.
(I) Core defense strategy: Conduct a no-arrest defense based on the absence of social danger, centering on Article 28.
This provision represents the core defense tool conferred upon defense counsel by the new regulations. The text enumerates seven mandatory circumstances under which arrest should not be authorized; these constitute rigid legal norms, such that, if a defendant meets the corresponding criteria, the procuratorial organ has no discretion and must issue a decision not to arrest. When handling a case, lawyers should promptly collect written evidence pertaining to self‑confession, meritorious service, restitution and compensation, criminal reconciliation, first-time or occasional offenses, and accomplice status. In their legal opinion, they should precisely cite Article 28 and, item by item, demonstrate that the defendant poses no social danger, thereby justifying release on bail. At the same time, they should carefully examine whether the case file submitted by the public security authorities contains objective evidence establishing the existence of social danger. If no such documentary evidence or witness testimony supports such a finding, they should directly argue that the statutory conditions for arrest are not met, thereby significantly narrowing the prosecutor’s discretionary scope.
(II) Secondary defense strategy: Relying on Article 27 to argue for non-arrest due to insufficient evidence.
Article 27 of the new regulations lists eight circumstances in which insufficient evidence warrants a decision not to arrest. For cases involving broken chains of evidence, contradictions in key evidence, or reliance solely on isolated evidence, defense counsel may systematically identify all evidentiary defects and demonstrate that the existing evidence fails even to meet the minimum standard required for arrest, thereby directly seeking a decision not to arrest. It is important to note the double-edged sword effect in practice: Currently, most prosecutors, wary of Rule 24 on case review, tend to raise the evidentiary threshold for arrest to that required for prosecution. Consequently, in cases where the evidence clearly contains defects, they are more likely to directly decide not to authorize arrest, which, in the short term, benefits defense attorneys seeking to secure a decision not to arrest. However, the drawback is that once a prosecutor approves an arrest, it becomes significantly more difficult to argue for non-prosecution at the subsequent stage of reviewing and deciding whether to prosecute on the grounds of insufficient evidence.
(III) Advanced defense strategy: Cite Article 5 to explain the lawful litigation logic of "arrest first, then no prosecution."
For cases where the evidence only meets the threshold for arrest but is unlikely to be sufficiently supplemented within a short period to satisfy the prosecution standard, the defense counsel may, in a legal opinion, fully articulate the spirit of Article 5 and the eighteen-character case-handling logic summarized by the author, and explain to the handling prosecutor that there is an inherent hierarchical distinction between the arrest standard and the prosecution standard. Non-prosecution following arrest due to insufficient evidence is a legally permissible course of action under the relevant regulations. Furthermore, if Article 24 is subsequently amended to include tiered exculpatory provisions, such cases will not be deemed to constitute improper arrests. This approach can alleviate the prosecutor’s concerns about potential accountability and create room for a subsequent decision of non-prosecution for the client.
(IV) Document Optimization: Abandon generic templates and develop specialized legal opinions with binding force.
In the past, criminal defense lawyers often relied on generic templates for legal opinions arguing against arrest, which were often vague in content and lacked thorough reasoning, making it difficult to persuade prosecutors to substantively adopt their views.
Following the implementation of the new regulations, lawyers should strategically refine the logical structure of their written submissions as follows: First, precisely cite the statutory provisions corresponding to the 2026 standards for the quality and efficiency of arrest reviews, including Articles 5, 7, 27, 28, and 24, ensuring that each statutory reference is directly aligned with the relevant facts of the case. Second, systematically deconstruct the three key requirements for arrest—evidence, sentencing considerations, and social dangerousness—by grounding arguments in the objective evidence contained in the case file and avoiding subjective or emotionally charged language. Third, clearly enumerate the normative and policy bases, as well as relevant precedent cases, supporting each defense position, so that the submission is complete, verifiable, and can be filed as part of the case record, thereby endowing the defense arguments with traceable, assessable, and legally binding force and increasing the likelihood of their acceptance by the procuratorial authorities.
6. Conclusion
The issuance of the 2026 standards for reviewing and approving arrests represents a significant institutional innovation by the procuratorial organs in implementing the criminal justice policy of fewer arrests, cautious prosecution, and prudent detention, as well as in advancing the standardized governance of pretrial detention. Article 5, in particular, clearly distinguishes between the two independent review standards for arrest and prosecution, setting forth the fundamental requirement that these standards must not be conflated. This provision marks a milestone in rectifying the long-standing judicial practice of “arresting upon establishment of a crime and prosecuting upon arrest,” reducing unnecessary pre-trial detention, and safeguarding the basic personal liberty of criminal suspects.
However, the review rules for improperly handled arrest cases under Article 24 contain drafting flaws: they fail to distinguish among the various reasons for non-prosecution and do not establish clear boundaries for exemption from liability. This creates a conflict with the spirit of Article 5, and in practice renders the new rule’s intended procedural pathway—arrest based on evidence followed by non-prosecution—difficult to implement, thereby undermining the overall effectiveness of the new regulation.
To resolve the inherent conflict between the two, it is not sufficient to simply negate the supervisory value of quality assessments. Rather, Article 24 should be refined and amended by introducing tiered evaluation criteria and compliance-based exculpatory provisions, accompanied by the issuance of detailed guidelines for their application, thereby reconstructing a diversified assessment mechanism that balances case-handling supervision with prosecutorial discretion.
At the same time, as criminal defense attorneys, we should proactively gain a thorough understanding of the complete statutory framework of the new regulations, firmly grasp key defense provisions such as Article 5, Article 27, and Article 28, and move away from a crude, template‑based, experience‑driven approach to defense. Instead, we should adopt a rules‑oriented, evidence‑supported, and meticulously crafted substantive defense strategy, precisely apply the relevant statutory provisions to offer professional legal opinions, fully safeguard our clients’ legitimate procedural rights, and promote the integration of legal, social, and human rights‑protection outcomes in the review‑and‑arrest process.
Zhuosheng Criminal
Team Introduction
Zhiheng · Introduction