🔷 Regulation of Algorithmic Accountability in Biotechnology in China

1. Introduction and Legal Framework

Algorithmic accountability in biotechnology refers to the legal responsibility of organizations that develop or use algorithms and artificial intelligence (AI) in genetic testing, genomic analysis, drug discovery, clinical diagnosis, precision medicine, and biotechnology research. It requires that algorithmic systems be sufficiently safe, transparent, reliable, and subject to human oversight.

China does not have one comprehensive statute exclusively governing AI algorithms used in biotechnology. Instead, regulation arises from several overlapping legal instruments:

Cybersecurity Law (2016, as amended): Regulates network security and related obligations.

Data Security Law (2021): Establishes data-security responsibilities and risk-management requirements.

Personal Information Protection Law (PIPL, 2021): Regulates personal information processing, including sensitive medical and genetic information.

Biosecurity Law (2021): Addresses biotechnology-related risks and biological security.

Regulations on the Administration of Human Genetic Resources (2019, amended in 2024): Govern the collection, preservation, utilization, and external provision of human genetic resources.

Provisions on the Administration of Algorithmic Recommendations in Internet Information Services (2022): Regulate specified algorithmic recommendation services.

Interim Measures for the Management of Generative Artificial Intelligence Services (2023): Apply to covered generative AI services provided to the public in China.

Regulations on the Administration of Network Data Security (2024, effective 1 January 2025): Supplement China’s network-data security framework.

The application of each instrument depends on the activity involved. For example, an algorithm used internally to analyze genetic sequences may be regulated differently from a public-facing AI medical consultation service.

2. Algorithmic Transparency and Accountability under the PIPL

Article 24 of the PIPL is particularly important for algorithmic accountability. It requires transparency and fairness in automated decision-making and prohibits unreasonable differential treatment in transaction conditions based on automated decisions.

Where automated decision-making has a significant impact on an individual's rights and interests, the individual has the right to request an explanation and to refuse a decision made solely through automated decision-making, subject to the law's scope and conditions.

In biotechnology, these requirements may apply to algorithms that:

Assess genetic susceptibility to diseases.

Rank patients for access to clinical trials.

Generate diagnostic or treatment recommendations.

Determine insurance or healthcare eligibility using genetic information.

Analyze personal medical records to predict disease risks.

Organizations should validate models, document their limitations, monitor discriminatory outcomes, and provide meaningful human review where appropriate. Genetic information is particularly sensitive because algorithmic errors can affect patients and their relatives.

3. Regulation of AI and Biotechnology Algorithms

China's algorithm-specific rules apply according to the nature of the service.

The Algorithmic Recommendation Provisions (2022) establish obligations for covered internet information services, including algorithmic transparency and user protections. They do not automatically govern every scientific algorithm used in a laboratory.

The Generative AI Interim Measures (2023) apply to covered generative AI services provided to the public in China. Their requirements may become relevant when biotechnology companies offer public-facing AI systems for medical information or health-related assistance. Internal research systems are not automatically subject to every provision of these measures.

Medical AI products may also fall under the medical-device regulatory framework administered by the National Medical Products Administration (NMPA). Where an algorithm is incorporated into a regulated medical device, applicable registration, clinical evaluation, quality-management, and post-market obligations must be considered.

4. Relevant Cases and Judicial Examples

He Jiankui criminal case (2019): The Shenzhen Nanshan District People's Court sentenced He Jiankui to three years' imprisonment and fined him RMB 3 million for illegally conducting human-embryo gene-editing activities involving reproductive medicine. The case was not about algorithmic accountability, but it demonstrates that scientific innovation does not exempt biotechnology researchers from legal and ethical obligations.

Didi Global cybersecurity enforcement (2022): China's cybersecurity regulator imposed a RMB 8.026 billion penalty on Didi Global for serious violations of personal-information and data-security laws. Although not a biotechnology or algorithmic-diagnosis case, it illustrates the consequences of failures in data governance.

Direct biotechnology AI precedent: A widely established Chinese judicial precedent specifically deciding liability for a faulty genomic prediction algorithm or an AI-based medical diagnosis remains difficult to identify. The above matters should therefore be treated as related enforcement examples, not direct precedents establishing AI-specific liability.

5. Liability and Compliance Measures

Depending on the violation, biotechnology organizations may face administrative penalties, orders to correct unlawful processing, suspension of relevant services, civil compensation claims, or criminal liability where the statutory requirements are satisfied.

A sound compliance programme should include algorithm validation, documented training-data provenance, genetic-data access controls, bias testing, cybersecurity safeguards, human oversight, explainability appropriate to the use case, and procedures for investigating adverse outcomes.

Conclusion: China's approach to algorithmic accountability in biotechnology is based on overlapping data-protection, cybersecurity, biosecurity, AI-service, and medical-device rules. The PIPL provides particularly important protections for individuals affected by automated decisions. However, legal obligations depend on the algorithm's function, the sensitivity of the data, and whether the system is used for research, clinical care, or a regulated public-facing service.