个性化推荐算法的法律风险规制

Legal Risk Regulation of Personalized Recommendation Algorithms

信息爆发增长催生了个性化推荐算法技术的兴起。个性化推荐算法在解决信息过载和长尾问题、满足用户个性化需求、提高互联网信息服务效率的同时,也引发了用户意思自治受限、隐私泄露、信息茧房、算法歧视等诸多法律风险,亟需法律作出必要的回应。为此,应当在诚信原则、自主原则、公正原则、比例原则的指导下,树立开放的隐私保护观,强化算法告知义务与用户拒绝权利,完善算法解释权,构建算法审计制度,以降低个性化推荐算法所带来的法律风险。

Information overload has led to the rise of personalized recommendation algorithm for information distribution.Personalized recommendation algorithms,while solving problems of information overload and long tail,meeting users’personalized needs,and improving the efficiency of Internet information service,have also triggered many legal risks such as the interference with with users’self-generated meaning,privacy leakage,information cocoon,and algorithm discrimination.The mission of jurisprudence is not to praise the achievements of personalized recommendation algorithms,but to examine their potentially irrational consequences,and to explore how to reduce the risks possibly caused by the development of algorithmic recommendations through the rule of law.To this end,a responsive algorithm regulation should establish an open view of privacy protection,strengthen the obligation of algorithm notification and the right of algorithm application rejection,improve the right of algorithm interpretation,and build an algorithm audit system to reduce the legal risks brought by personalized recommendation algorithms under principles of integrity,autonomy,justice and proportionality.