This paper shows that the law, in subtle ways, may set hitherto unrecognized incentives for the adoption of explainable machine learning applications. In doing so, we make two novel contributions. First, on the legal side, we show that to avoid liability, professional actors, such as doctors and managers, may soon be legally compelled to use explainable ML models. We argue that the importance of explainability reaches far beyond data protection law, and crucially influences questions of contractual and tort liability for the use of ML models. To this effect, we conduct two legal case studies, in medical and corporate merger applications of ML. As a second contribution, we discuss the (legally required) trade-off between accuracy and explainability and demonstrate the effect in a technical case study in the context of spam classification.

/pdf/explainable-ai-under-contract-and-tort-law-legal-incentives-34m9c3rufr.pdf

Explainable AI under contract and tort law: legal incentives and technical challenges

Analyzing recommender systems for health promotion using a multidisciplinary taxonomy: A scoping review.

Demand is growing for more accountability regarding the technological systems that increasingly occupy our world. However, the complexity of many of these systems—often systems-of-systems—poses accountability challenges. A key reason for this is because the details and nature of the information flows that interconnect and drive systems, which often occur across technical and organizational boundaries, tend to be invisible or opaque. This paper argues that data provenance methods show much promise as a technical means for increasing the transparency of these interconnected systems. Specifically, given the concerns regarding ever-increasing levels of automated and algorithmic decision-making, and so-called “algorithmic systems” in general, we propose decision provenance as a concept showing much promise. Decision provenance entails using provenance methods to provide information exposing decision pipelines: chains of inputs to, the nature of, and the flow-on effects from the decisions and actions taken (at design and run-time) throughout systems. This paper introduces the concept of decision provenance, and takes an interdisciplinary (tech-legal) exploration into its potential for assisting accountability in algorithmic systems. We argue that decision provenance can help facilitate oversight, audit, compliance, risk mitigation, and user empowerment, and we also indicate the implementation considerations and areas for research necessary for realizing its vision. More generally, we make the case that considerations of data flow, and systems more broadly, are important to discussions of accountability, and complement the considerable attention already given to algorithmic specifics.

Decision Provenance: Harnessing Data Flow for Accountable Systems

Using artificial intelligence (AI) technology to replace human decision-making will inevitably create new risks whose consequences are unforeseeable. This naturally leads to calls for regulation, b...

https://royalsocietypublishing.org/doi/pdf/10.1098/rsta.2017.0360

How should we regulate artificial intelligence

A survey on multi-factor authentication for online banking in the wild

This paper investigates the question of legal liability for the consequences of decisions made by machine learning technology rather than by humans, although we do not attempt a detailed analysis of the basis on which such liability might be imposed. This is a substantial task which would require far more space than is available here.The initial focus is on private claims for personal injury, property damage and other losses caused by use of machine learning technologies. These claims will usually be made via the tort of negligence.Equally importantly, we identify some of the threats to individual autonomy and fundamental rights which are created by the use of machine learning to make decisions. Breach of those fundamental rights is a second source of potential liability.We conclude by suggesting a potential link between liability and the preservation of those fundamental rights which might achieve an interim solution to this issue, making use of the concept of accountability and its transparency attribute in particular.

Responsibility, Autonomy and Accountability: Legal Liability for Machine Learning

Data security and multi-factor authentication: Analysis of requirements under EU law and in selected EU Member States

This report (the “Report”) considers certain legal requirements relating to data security in the EU, and specifically the use of multi-factor authentication as a method of meeting the security obligations established by European Directive 95/46 EC on the processing of personal data (the “Directive”). Following this Executive Summary, the Report comprises two sections: a discussion of the requirements of data security under European data protection legislation, and a study of selected national positions.

Data Security and Multi-Factor Authentication: Analysis of Requirements Under EU Law and in Selected EU Member States

Debates about the relationship between race and class, racism and capitalism, have been with us for as long as there’s been a socialist movement. In the past few years they have surfaced in and around the Democratic Socialists of America (DSA), particularly focused on the controversial writings of Adolph Reed. Reed was scheduled to give a talk sponsored by DSA’s Philadelphia and Lower Manhattan branches in late May, but he canceled his appearance after facing criticism from DSA’s Afrosocialists and Socialists of Color Caucus.

Elizabeth J Kennedy

Papers

Responsibility, Autonomy and Accountability: Legal Liability for Machine Learning

Data security and multi-factor authentication: Analysis of requirements under EU law and in selected EU Member States

Data Security and Multi-Factor Authentication: Analysis of Requirements Under EU Law and in Selected EU Member States

Racism and Capitalism