AI and the Ethics of Automating Consent

AI systems collect, process, and generate data in ways that further exacerbate many long-documented problems with online consent, most notably issues of providing adequate notice, choice, and withdrawal to users. The unpredictable and even unimaginable use of data by AI systems is considered a feature, not a bug. Yet this feature creates problems for notifying users as well as assessing when consent might be required based on potential uses, harms, and consequences. This article investigates whether these problems impact morally transformative consent in AI systems. We argue that while supplementing consent with further mechanization, digitization, and intelligence—either through proffering notification on behalf of the consentee or choosing and communicating consent by the consenter—may improve take-it-or-leave-it notice and choice consent regimes, the goal for AI consent should be one of partnership development between parties, built on responsive design and continual consent.

[1]  Daniel J. Solove,et al.  Privacy Self-Management and the Consent Dilemma , 2012 .

[2]  Bart W. Schermer,et al.  The crisis of consent: how stronger legal protection may lead to weaker consent in data protection , 2014, Ethics and Information Technology.

[3]  Helen Nissenbaum,et al.  Big data's end run around procedural privacy protections , 2014, Commun. ACM.

[4]  Tribhuvanesh Orekondy,et al.  Towards a Visual Privacy Advisor: Understanding and Predicting Privacy Risks in Images , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[5]  Lujo Bauer,et al.  Privacy Expectations and Preferences in an IoT World , 2017, SOUPS.

[6]  Bin Liu,et al.  Automated Analysis of Privacy Requirements for Mobile Apps , 2016, NDSS.

[7]  M. Calo.  Against Notice Skepticism In Privacy (And Elsewhere) , 2011 .

[8]  Alfred Kobsa,et al.  Towards Privacy-Aware Smart Buildings: Capturing, Communicating, and Enforcing Privacy Policies and Preferences , 2017, 2017 IEEE 37th International Conference on Distributed Computing Systems Workshops (ICDCSW).

[9]  Lorrie Faith Cranor,et al.  A Design Space for Effective Privacy Notices , 2015, SOUPS.

[10]  David A. Wagner,et al.  The Feasibility of Dynamically Granted Permissions: Aligning Mobile Privacy with User Preferences , 2017, 2017 IEEE Symposium on Security and Privacy (SP).

[11]  Tristan Henderson,et al.  Contextual Consent: Ethical Mining of Social Media for Health Research , 2017, ArXiv.

[12]  Mario Pascalev,et al.  Privacy exchanges: restoring consent in privacy self-management , 2016, Ethics and Information Technology.

[13]  Lorrie Faith Cranor,et al.  Necessary But Not Sufficient: Standardized Mechanisms for Privacy Notice and Choice , 2012, J. Telecommun. High Technol. Law.

[14]  Alessandro Acquisti,et al.  Follow My Recommendations: A Personalized Privacy Assistant for Mobile App Permissions , 2016, SOUPS.

[15]  Bernt Schiele,et al.  Speaking the Same Language: Matching Machine to Human Captions by Adversarial Training , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).