Provisions on the Management of Algorithmic Recommendations - AI Regulation Reference

Provisions on the Management of Algorithmic Recommendations

Jurisdiction:

China

enforcing

Effective:

Mar 1, 2022

Authority:

Cyberspace Administration of China

Official text Verified Mar 26, 2026

Obligations Covered

Transparency & Disclosure Risk Assessment

Provisions (2)

Algorithm Filing and Registration #

Obligation:

enforcing

Effective:

Mar 1, 2022

Risk tier:

all

Scope:

providers

Requirements

Requirement	Details
Algorithm filing	Providers must file algorithm details with regulators
Public disclosure	Must disclose basic principles of algorithmic recommendations to users
User controls	Users must be able to turn off algorithmic recommendations
Content management	Providers must review and manage content recommended by algorithms

Penalties

Violation	Fine
Non-compliance	Warnings, corrections, service suspension per Cybersecurity Law

Algorithmic Risk Assessment #

Obligation:

Risk Assessment

enforcing

Effective:

Mar 1, 2022

Risk tier:

all

Scope:

providers

Requirements

Requirement	Details
Security assessment	Services with public opinion properties require security assessment
Content review	Must establish review mechanisms for algorithmic models and training data
Prohibited uses	Cannot use algorithms to engage in activities endangering national security
Monitoring	Continuous monitoring for illegal content in algorithmic outputs

Penalties

Violation	Fine
Non-compliance	Warnings, corrections, service suspension per Cybersecurity Law