人工智能时代算法歧视风险及治理研究
Algorithmic Discrimination and Its Governance in Era of Artificial Intelligence

人工智能时代，在人工智能与大数据技术深度融合的时代背景下，算法借助各类数据实现倍增式的发展，带来便利的背后也引起不同程度的算法歧视。算法歧视作为传统社会歧视的数字化延伸，算法歧视的成因包括：算法黑箱带来的算法不可解释和决策不透明、数据偏差对文化历史偏见的继承放大、设计者和使用者的主观思维嵌入等。它的形成不仅加剧数字时代的信息失衡，抑制了创新发展的内生动力，更对社会的伦理秩序与法治根基构成深层冲击。规制算法歧视需多方协同发力，量化算法歧视风险程度，细化相关法律条款，构建更为细致明确的责任划分体系，设计者和使用者提高数字素养和算法认知、企业和社会需对算法的运用和监管勠力同心，构建责任共担的治理机制，多方协同护航数字社会的公平正义与可持续发展。
In the era of artificial intelligence, against the backdrop of the deep integration of artificial intelligence and big data technologies, algorithms have achieved exponential growth through leveraging diverse datasets. While bringing convenience, they have also engendered varying degrees of algorithmic discrimination. As a digital extension of traditional social discrimination, the causes of algorithmic discrimination include: The unexplainability of algorithmic operations and decision-making opacity stemming from algorithmic black boxes; the perpetuation and amplification of cultural-historical prejudices through data biases, and subjective thinking embedding of by designers and users. Its emergence not only exacerbates information imbalance in the digital age and stifles endogenous drivers for innovative development, but also poses profound challenges to the ethical order and legal foundation of society. Regulating algorithmic discrimination requires the concerted efforts of multiple parties, quantifying the risk level of algorithmic discrimination, refining relevant legal provisions, establishing a more detailed and clear responsibility allocation system, enhancing the digital literacy and algorithmic awareness of designers and users, and enterprises and society need to work together to apply and supervise algorithms, building a governance mechanism of shared responsibility to jointly safeguard the fairness, justice and sustainable development of the digital society.