20220214-世界经济论坛-Advancing_Digital_Agency_The_Power_of_Data_Intermediaries_46页_2mb

Advancing Digital Agency: The Power of Data Intermediaries

This report explores the role of data intermediaries in advancing digital agency and addressing the challenges of navigating complex data ecosystems, particularly issues of trust, control, and human-centricity. Core challenges include mistrust in data sharing, complex value chains, and excessive reliance on nuanced digital consent mechanisms that hinder user agency.

Key insights include:

• Data intermediaries unlock potential to improve trust, data portability, and the delegation of user preferences by acting as neutral, trustworthy third parties.
• Data value chains, as explored via frameworks like Open Data Watch, highlight opportunities for intermediaries to substitute person-centric control with flexible permission-free mechanisms.
• Human-centric agency can be preserved by enabling the delegation of agency via trusted data intermediaries, particularly digital agents (TDAs), as discussed in the proposed “Valexander” model.

Three major organizational models for intermediaries were analyzed: data stewards, data trusts, and data collaboratives. Another model discussed was the digital fiduciary, bound by duties of care and loyalty. Support from robust frameworks including digital identity, governance standards, liability management, and technical solutions (APIs, Zero-Knowledge proofs) are crucial.

Data governance policymaking must confront issues such as defining neutrality for data intermediaries (as per GDPR), securing cross-border data transfers, and establishing standards for regulation. Public-private partnerships are recommended to harmonize innovations.

The future of digital agency hinges on trusted intermediaries and robust policy frameworks. Regulatory sandbox approaches offer promising pathways for testing data sharing innovations. Moving forward requires a multi-stakeholder call to action addressing the challenges of trust in the face of complex automated decision-making systems, particularly in AI governance.