智能城市的算法透明度

Algorithmic Transparency for Smart Cities

公共机构将数据分析交由私人供应商负责,使得机构代理人员与公众均无法获知算法决策的设计及其背后的政策选择。通过向23个州政府提交42份公开记录请求,寻求有关6个预测算法决策的基本信息,以探究公开记录流程能否让公民发现这些算法所体现的政策判断,并评估其效用和公平性,但这些请求基本上未得到回复。传统观念认为,过度宽泛的商业保密保护是无法产生算法透明度的主要原因,但该观念并不准确。只有在以下情况下,公共部署算法才会足够透明:1.政府在算法运行和后续实施以及验证目标时产生的适当记录;2.私人供应商向公共机构披露有关它们如何开发算法的充分信息;3.公共机构和法院将商业秘密主张视为法律要求的公开披露的有限例外。

Public agencies entrust data analysis to private vendors,making it impossible for public agents and the public to know the design of algorithm decisions and the policy choices behind them.By submitting 42 public record requests to 23 state governments to seek basic information about the decisions made by the six prediction algorithms,to assess whether the public recording process enables citizens to discover the policy judgments embodied in these algorithms,and to assess their effectiveness and fairness,but these requests went largely unanswered.The traditional view is that the over-broad protection of commercial secrecy is the main reason for the lack of algorithm transparency,but this concept is not accurate.Publicly deployed algorithms will be sufficiently transparent only if(1)governments generate appropriate records about their objectives for algorithmic processes and subsequent implementation and validation;(2)government contractors reveal to the public agency sufficient information about how they developed the algorithm;and(3)public agencies and courts treat trade secret claims as the limited exception to public disclosure that the law requires.