生成式人工智能药物研发数据权利困境与治理.pdfVIP

专论

医学与哲学2025

年7

月第46

卷第

13

期总第768期

生成式人工智能药物研发数据权利困境与治理

庞钰1

摘要：生成式人工智能以数据密集型范式高效整合跨地域患者和基因数据，重塑了药物研发流程，

但也带来了数据控制削弱、权利主体弱化与数据公平缺位等数据权利困境。在此基础上，提出

“可解释−可协商−可追责”三维数据权利治理框架，强调通过分层解释与公共平台降低信息壁垒；

用事件驱动再授权和多方协商保障数据主体持续参与；以责任节点标注、问责委员会和风险预警

闭合责任链来实现人−机−数据共治并重新确认数据主体的道德地位，在加速药物创新的同时维护

个体尊严与社会公正。

关键词：生成式人工智能，药物研发，数据权利，伦理困境，算法治理

中图分类号：R-05文献标识码：A文章编号：1002-0772(2025)13-0012-05

DOI：10.12014/j.issn.1002-0772.2025.13.03

1

TheDilemmaandGovernanceofDataRightsinGenerativeArtificialIntelligence–drivenDrugDevelopmentPANGYu．

1.

SchoolofGovernment,PekingUniversity,Beijing100871,China

Abstract：Generative

artificial

intelligence

reshapes

drug

development

with

a

data-intensive

paradigm,

efficiently

integrates

cross-regional

patient

and

genetic

data,

but

also

brings

data

rights

dilemmas

such

as

weakened

data

control,

weakened

rights

subjects,

and

a

lack

of

data

equity.

On

this

basis,

a

three-dimensional

data

rights

governance

framework

of

explainable-

negotiable-accountable

is

proposed,

emphasizing

the

reduction

of

information

barriers

through

layered

interpretation

and

public

platforms.

The

study

promotes

sustained

data

subject

engagement

via

event-driven

reauthorization

and

multi-stakeholder

negotiation,

and

ensures

human–machine–data

co-governance

thr