Abstract
Data sharing poses complex ethical questions for data management. Manifold conflicting and shifting values need to be reconciled in pursuing viable data-management policies. For example, how does one make data available in useable form to stakeholders including scientists, governments and businesses worldwide, while assuring confidentiality, satisfying one's research ethics committee, protecting intellectual property and national security, and containing costs? Increasingly, ethical problem solving requires integration of ethics with technological "know how" and empirical research on the presenting problem. Each problem is highly contextual; broad application of general ethical principles such as always practice openness, or prepare all data for sharing, may have harmful unintended consequences. Chaos theory provides a heuristic or vision for understanding and coping with complexity and uncertainty. It does not provide answers to problems of data management, but frames the issues, and provides appropriate expectations and heuristics for considering data management problems.