ABSTRACT. Value alignment is a property of an intelligent agent indicating that it can only pursue goals and activities that are beneficial to humans. Traditional approaches to value alignment use imitation learning or preference learning to infer the values of humans by observing their behavior. We introduce a complementary technique in which a value-aligned prior is learned from naturally occurring stories which encode societal norms. Training data is sourced from the children's educational comic strip, Goofus & Gallant. In this work, we train multiple machine learning models to classify natural language descriptions of situations found in the comic strip as normative or non-normative by identifying if they align with behaviors of the main characters. We also report the models' performance when transferring to two unrelated tasks with little to no additional training on the new task.

ABSTRACT. In future agent societies, we might see AI systems engaging in selfish, calculated behavior, furthering their owners' interests instead of socially desirable outcomes. How can we promote morally sound behaviour in such settings, in order to obtain more desirable outcomes? A solution from moral philosophy is the concept of a \emph{social contract}, a set of rules that people would voluntarily commit to in order to obtain better outcomes than those brought by anarchy. We adapt this concept to a game-theoretic setting, to systematically modify the payoffs of a non-cooperative game, so that agents will rationally pursue socially desirable outcomes.

We show that for any game, a suitable social contract can be designed to produce an optimal outcome in terms of social welfare. We then investigate the limitations of applying this approach to alternative moral objectives, and establish that, for any alternative moral objective that is significantly different from social welfare, there are games for which no such social contract will be feasible that produces non-negligible social benefit compared to collective selfish behaviour.

ABSTRACT. On a variety of complex decision-making tasks, from doctors prescribing treatment to judges setting bail, machine learning algorithms have been shown to outperform expert human judgments. One complication, however, is that it is often difficult to anticipate the effects of algorithmic policies prior to deployment, as one generally cannot use historical data to directly observe what would have happened had the actions recommended by the algorithm been taken. A common strategy is to model potential outcomes for alternative decisions assuming that there are no unmeasured confounders (i.e., to assume ignorability). But if this ignorability assumption is violated, the predicted and actual effects of an algorithmic policy can diverge sharply. In this paper we present a flexible Bayesian approach to gauge the sensitivity of predicted policy outcomes to unmeasured confounders. In particular, and in contrast to past work, our modeling framework easily enables confounders to vary with the observed covariates. We demonstrate the efficacy of our method on a large dataset of judicial actions, in which one must decide whether defendants awaiting trial should be required to pay bail or can be released without payment.

ABSTRACT. Although essential to revealing biased performance, well intentioned attempts at algorithmic auditing can have effects that may harm the very populations these measures are meant to protect. This concern is even more salient while auditing biometric systems such as facial recognition, where the data is sensitive and the technology is often used in ethically questionable manners. We demonstrate a set of five ethical concerns in the particular case of auditing commercial facial processing technology, highlighting additional design considerations and ethical tensions the auditor needs to be aware of so as not exacerbate or complement the harms propagated by the audited system. We go further to provide tangible illustrations of these concerns, and conclude by reflecting on what these concerns mean for the role of the algorithmic audit and the fundamental product limitations they reveal.

ABSTRACT. Inspired by recent breakthroughs in predictive modeling, practitioners in both industry and government have turned to machine learning with hopes of operationalizing these predictions to drive decisions. Unfortunately, many social desiderata concerning consequential decisions, such as justice or fairness, have no natural formulation within a purely predictive framework. In the hopes of mitigating these problems, researchers have proposed a variety of metrics for quantifying various notions of deviation from the parities we might observe in a perfect world and offered a variety of algorithms that attempt to satisfy subsets of these parities or to trade off the degree to which they are satisfied against utility. In this paper, we connect this approach to the literature on ideal and non-ideal methodological approaches in political philosophy. The ideal approach consists of positing a perfect world, assessing deviations between our world and the perfect world, and taking actions to minimize these discrepancies wherever we observe them. However, by failing to account for the mechanisms by which our non-ideal world arose, the responsibilities of various decision-makers, and the impacts of their actions, ideal thinking can often lead to misguided policies. In this paper we demonstrate a connection between the recent literature on fair machine learning and the ideal approach in political philosophy, and show that some recently uncovered shortcomings in proposed algorithms reflect broader troubles faced by the ideal approach. We work this analysis through for both statistical and causal formulations of fairness and suggest several directions for new research.

ABSTRACT. Modern face recognition systems leverage datasets containing images of hundreds of thousands of specific individuals' faces to train deep convolutional neural networks to learn an embedding space that maps an arbitrary individual's face to a vector representation of their identity. The performance of a face recognition system in face verification (1:1) and face identification (1:N) tasks is directly related to the ability of an embedding space to discriminate between identities. Recently, there has been significant public scrutiny into the source and privacy implications of large-scale face recognition training datasets such as MS-Celeb-1M and MegaFace, as many people are uncomfortable with the idea of their face being used to train dual-use technologies such as face recognition systems. However, the actual impact of an individual's inclusion in such a dataset on a derived system's ability to recognize them has not previously been studied. In this work, we audit ArcFace, a state-of-the-art, open source face recognition system, in a large-scale face identification experiment with more than one million distractor images. We find a rank-1 face identification accuracy of 79.71% for individuals present in the model's training data and an accuracy of 75.73% for those not present. This modest difference in accuracy demonstrates that modern face recognition systems are biased towards individuals they are trained on, which has serious privacy implications when one considers that all major open-source face recognition training datasets do not gather informed consent from individuals during their collection.

ABSTRACT. AI is increasingly being offered 'as a service' (AIaaS). This entails service providers offering customers access to pre-built models, for tasks such as object recognition, text translation, text-to-voice conversion, and facial recognition, to name a few. The offerings enable customers to easily integrate a range of powerful ML-driven capabilities into their applications. Customers access these models through the provider's APIs, sending particular data to which the model is applied, and results returned.

However, there are many situations in which the use of ML can be problematic. AIaaS services typically represent generic functionality, available to customers at 'a few clicks'. Providers may therefore, for reasons of reputation or responsibility, seek to ensure that the AIaaS services they offer are being used by customers for 'appropriate' purposes.

This paper introduces and explores a concept in which AIaaS providers uncover situations of possible service misuse by their customers. Illustrated through topical examples, we consider the technical usage patterns that could signal situations warranting scrutiny, and raise some of the legal and technical challenges of monitoring for misuse. In all, by introducing this concept, we indicate a potential area for further inquiry from a range of perspectives.

ABSTRACT. Algorithmic (including AI/ML) decision-making artifacts are an established and growing part of our decision-making ecosystem. They have become near-indispensable as tools to help manage the flood of information we need to make timely effective decisions in an increasingly complex world. The current literature is awash with examples of how individual artifacts violate societal norms and expectations (e.g. violations of fairness, privacy, or safety norms). Against this backdrop, we highlight the need for principled frameworks for assessing value misalignment in AI-equipped sociotechnical systems. One trend in research explorations of value misalignment in artifacts is the focus on the behavior of singular tech artifacts. In this discussion, we outline and argue for a more structured systems-level approach for assessing value-alignment in sociotechnical systems. The discussion focuses primarily on fairness audits. We use the opportunity to highlight how adopting a system perspective improves our ability to explain and address value misalignments better. Our discussion ends with an exploration of priority questions that demand attention if we are to assure the value alignment of whole systems, not just individual artifacts.

ABSTRACT. Many researchers work on improving the data efficiency of machine learning. What would happen if they succeed? This paper explores the social-economic impact of increased data efficiency. Specifically, we examine the intuition that data efficiency will erode the barriers to entry protecting incumbent data-rich AI firms, exposing them to more competition from data-poor firms. We find that this intuition is only partially correct: data efficiency makes it easier to create ML applications, but large AI firms may have more to gain from higher performing AI systems. Further, we find that the effect on privacy, data markets, robustness, and misuse are complex. For example, while it seems intuitive that misuse risk would increase along with data efficiency -- as more actors gain access to any level of capability -- the net effect crucially depends on how much defensive measures are improved. More investigation into data efficiency, as well as research into the ``AI production function", will be key to understanding the development of the AI industry and its societal impacts.

ABSTRACT. As AI systems become prevalent in high stakes domains such as surveillance and healthcare, researchers now examine how to design and implement them in a safe manner. However, the potential harms caused by systems to stakeholders in complex social contexts and how to address these remains unclear. In this paper, we explain the inherent normative uncertainty in debates about the safety of AI systems. We then address this as a problem of vagueness by examining its place in the design, training, and deployment stages of AI system development. We adopt Ruth Chang's theory of intuitive comparability to illustrate the dilemmas that manifest at each stage. We then discuss how stakeholders can navigate these dilemmas by incorporating distinct forms of dissent into the development pipeline, drawing on Elizabeth Anderson's work on the epistemic powers of democratic institutions. We outline a framework of sociotechnical commitments to formal, substantive and discursive challenges that address normative uncertainty across stakeholders, and propose the cultivation of related virtues by those responsible for development.

ABSTRACT. In this paper, we present the Algorithmic Impact Assessment (AIA) of personalized wellbeing recommendations delivered through Telefónica Alpha’s app REM!X. The main goal of the AIA was to identify potential algorithmic biases in the recommendations that could lead to the discrimination of protected groups. The assessment was conducted through a qualitative methodology that included five focus groups with developers and a digital ethnography relying on users comments reported in the Google Play Store. To minimize the collection of personal information, as required by best practice and the GDPR, the REM!X app did not collect gender, age, race, religion, or other protected attributes from its users. This limited the algorithmic assessment and the ability to control for different algorithmic biases. Nevertheless, based on indirect evidence, the AIA identified four hypothetical domains that put the levels of fairness and discrimination at risk. Our analysis provided important insights about the impact of color blindness on algorithmic audit and transparency, and how to address it

ABSTRACT. Reward functions are easy to misspecify; although designers can make corrections after observing mistakes, an agent pursuing a misspecified reward function can irreversibly change the state of its environment. If that change precludes optimization of the correctly specified reward function, then correction is futile. For example, a robotic factory assistant could break expensive equipment due to a reward misspecification; even if the designers immediately correct the reward function, the damage is done. To mitigate this risk, we introduce an approach that balances optimization of the primary reward function with preservation of the ability to optimize auxiliary reward functions. Surprisingly, even when the auxiliary reward functions are randomly generated and therefore uninformative about the correctly specified reward function, this approach induces conservative, effective behavior.

ABSTRACT. Data sharing for medical research has been difficult as open-sourcing clinical data may violate patient privacy. Creating openly available datasets on medical videos, especially videos where faces are necessary for diagnosis, is infeasible unless the ethical requirements are met. Traditional face de-identification methods wipe out facial information entirely, making it impossible to analyze facial behavior. Recent advancements on whole-body keypoints detection also rely on facial input to estimate body keypoints. Both facial and body keypoints are critical in some medical diagnoses, and keypoints invariability after de-identification is of great importance. Here, we propose a solution using deepfakes, the face swapping technique. While this swapping method has been criticized for invading privacy and portraiture right, it could conversely protect privacy in medical video: patients' faces could be swapped to a proper target face and become unrecognizable. However, it remains an open question that to what extent the swapping de-identification method affects the automatic detection of body keypoints. In this study, we apply deepfake technique to Parkinson's Disease examination videos to de-identify subjects, and quantitatively show that: face-swapping as a de-identification approach is reliable, and it keeps the keypoints almost invariant, significantly better than traditional methods. This study proposes a pipeline for video de-identification and keypoint preservation, clearing up ethical restrictions for medical data sharing. This work could make open source high quality medical video datasets more feasible and promote future medical research that benefits our society.

ABSTRACT. We propose a game-theoretical model to simulate the dynamics of AI adoption on scale-free networks with and without link rewiring. This formalism allows us to understand the impact of the adoption of AI systems for society as a whole, addressing some of the concerns on the need for regulation. Using this model we study the adoption of AI systems, the distribution of the different types of AI (from selfish to utilitarian), the appearance of clusters of specific AI types, and the impact on the fitness of each individual. We suggest that the entangled evolution of individual strategy and network structure constitutes a key mechanism for the sustainability of utilitarian and human-conscious AI. Differently, in the absence of rewiring, a minority of the population can easily foster the adoption of selfish AI and gains a benefit at the expense of the remaining majority.

ABSTRACT. This paper proposes the establishment of "Medical Artificial Intelligence (AI) Types (MA Types)" that classify AI in medicine not only by technical system requirements but also implications to healthcare workers’ roles and us-ers/patients. MA Types can be useful to promote discussion regarding the purpose and application of the clinical site. Although MA Types are based on the current technologies and regulations in Japan, but that does not hinder the potential reform of the technologies and regulations. MA Types aims to facilitate discussions among physicians, healthcare workers, engineers, public/patients and policymakers on AI systems in medical practices.

ABSTRACT. Graph clustering groups entities --- the vertices of a graph --- based on their similarity, typically using a complex distance function over a large number of features. Successful integration of clustering approaches in automated decision-support systems hinges on the interpretability of the resulting clusters. This paper addresses the problem of generating interpretable clusters, given features of interest that signify interpretability to an end-user, by optimizing interpretability in addition to common clustering objectives. We propose a $\beta-$interpretable clustering algorithm that ensures that at least $\beta$ fraction of nodes in each cluster share the same feature value. The tunable parameter $\beta$ is user-specified. We also present a more efficient algorithm for scenarios with $\beta=1$ and analyze the theoretical guarantees of the two algorithms. Finally, we empirically evaluate our approaches using four real-world datasets. The interpretability of the clusters is complemented by generating explanations in the form of labels denoting the feature values of the nodes in the clusters, using frequent pattern mining.

ABSTRACT. While natural language processing affords researchers an opportunity to automatically scan millions of social media posts, there is growing concern that automated computational tools lack the ability to understand context and nuance in human communication and language. This article introduces a critical systematic approach for extracting culture, context, and nuance in social media data. The Contextual Analysis of Social Media (CASM) approach considers and critiques the gap between inadequacies in natural language processing tools and differences in geographic, cultural, and age-related variance of social media use and communication. CASM utilizes a team-based approach to the analysis of social media data, explicitly informed by community expertise. We use of CASM to analyze Twitter posts from gang-involved youth in Chicago. We designed a set of experiments to evaluate the performance of a support vector machine us-ing CASM hand-labeled posts against a distant model. We found that the CASM-informed hand-labeled data outperforms the baseline distant labels, indicating that the CASM labels capture additional dimensions of information that content-only methods lack. We then question whether this is helpful or harmful for gun violence prevention.

ABSTRACT. The recent flood of concern around issues such as social biases implicit in algorithms, economic impacts of artificial intelligence (AI), and potential existential threats posed by the development of AI technology motivate consideration of regulatory action to forestall or constrain certain developments in the fields of AI and machine learning. However, definitional ambiguity hampers the possibility of conversation about these urgent topics of public concern. Legal and regulatory interventions require agreed-upon definitions, but consensus around a definition of AI has been elusive, especially in policy conversations. With an eye towards practical working definitions and a broader understanding of positions on these issues, we use a series of surveys and a review of published policy documents to examine variation in researcher and policy-maker conceptions of AI. We find that while AI researchers tend to favor definitions of AI that emphasize technical functionality, policy-makers favor definitions that emphasize comparison to human thinking and behavior. We point out that definitions that adhere closely to the functionality of AI systems are more inclusive of technologies in use today, whereas definitions that emphasize human-like capabilities are most applicable to hypothetical future technologies. As a result of this gap, ethical and regulatory efforts may emphasize concern about future technologies over pressing issues with existing deployed technologies.

ABSTRACT. Artificial intelligence (AI) has wide societal implications, yet social scientists are only beginning to study public attitudes toward the technology. Existing studies find that the public's trust in institutions can play a major role in shaping the regulation of emerging technologies. Using a large-scale survey (N=2000), we examined Americans' perceptions of 13 AI governance challenges as well as their trust in governmental, corporate, and multistakeholder institutions to responsibly develop and manage AI. While Americans perceive all of the AI governance issues to be important for tech companies and governments to manage, they have only low to moderate trust in these institutions to manage AI applications.

ABSTRACT. Certainty around the regulatory environment is crucial to facilitate responsible AI innovation and its social acceptance. However, the existing legal liability system is inapt to assign responsibility where a potentially harmful conduct and/or the harm itself are unforeseeable, yet some instantiations of AI and/or the harms they may trigger are not foreseeable in the legal sense. The unpredictability of how courts would handle such cases makes the risks involved in the investment and use of AI incalculable, creating an environment that is not conducive to innovation and may deprive society of some benefits AI could provide. To tackle this problem, we propose to draw insights from financial regulatory best-practices and establish a system of AI guarantee schemes. We envisage the system to form part of the broader market-structuring regulatory framework, with the primary function to provide a readily available, clear, and transparent funding mechanism to compensate claims that are either extremely hard or impossible to realize via conventional litigation. We propose at least partial industry-funding, with funding arrangements depending on whether it would pursue other potential policy goals.

ABSTRACT. Since 2016, more than 80 AI ethics documents – including codes, principles, frameworks, and policy strategies – have been produced by corporations, governments, and NGOs. In this paper, we examine three topics of importance related to our ongoing empirical study of ethics and policy issues in these emerging documents. First, we review possible challenges associated with the relative homogeneity of the documents’ creators. Second, we provide a novel typology of motivations to characterize both obvious and less obvious goals of the documents. Third, we discuss the varied impacts these documents may have on the AI governance landscape, including what factors are relevant to assessing whether a given document is likely to be successful in achieving its goals.

ABSTRACT. We present an innovative methodology for studying and teaching the impacts of AI through a role-play game. The game serves two primary purposes: 1) training AI developers and AI policy professionals to reflect on and prepare for future social and ethical challenges related to AI and 2) exploring possible futures involving AI technology development, deployment, social impacts, and governance. While the game currently focuses on the inter-relations between short-, mid- and long-term impacts of AI, it has potential to be adapted for a broad range of scenarios, exploring in greater depths issues of AI policy re-search and affording training within organizations. The game presented here has undergone two years of development and has been tested through over 30 events involving between 3 and 70 participants. The game is under active development, but preliminary findings suggest that role-play is a promising methodology for both exploring AI futures and training individuals and organizations in thinking about, and reflecting on, the impacts of AI and strategic mistakes that can be avoided today.

ABSTRACT. At the end of the Cold War, the renowned political scientist, Samuel Huntington, argued that future conflicts were more likely to stem from cultural frictions -- ideologies, social norms, and political systems -- rather than political or economic frictions. Huntington focused his concern on the future of geopolitics in a rapidly shrinking world. This paper argues that a similar dynamic is at play in the interaction of technology cultures. We emphasize the role of culture in the evolution of technology and identify the particular role culture (esp. privacy culture) plays in the development of AI/ML technologies. Then we examine some implications that this perspective brings to the fore.

ABSTRACT. There is growing concern over the potential misuse of artificial intelligence (AI) research. Publishing scientific research can facilitate misuse of the technology, but the research can also contribute to protections against misuse. This paper addresses the balance between these two effects. Our theoretical framework elucidates the factors governing whether the published research will be more useful for attackers or defenders, such as the possibility for adequate defensive measures, or the independent discovery of the knowledge outside of the scientific community. The balance will vary across scientific fields. However, we show that the existing conversation within AI has imported concepts and conclusions from prior debates within computer security over the disclosure of software vulnerabilities. While disclosure of software vulnerabilities often favours defence, this cannot be assumed for AI research. The AI research community should consider concepts and policies from a broad set of adjacent fields, and ultimately needs to craft policy well-suited to its particular challenges.

ABSTRACT. The artificial intelligence (AI) community has recently engaged in activism in relation to their employers, other members of the community, and their governments in order to shape the societal and ethical implications of AI. It has achieved some notable successes, but prospects for further political organising and activism are uncertain. We survey activism by the AI community over the last six years; apply two analytical frameworks drawing upon the literature on epistemic communities, and worker organising and bargaining; and explore what they imply for the future prospects of the AI community. Success thus far has hinged on a coherent shared culture, and high bargaining power due to the high demand for a limited supply of AI ‘talent’. Both are crucial to the future of AI activism and worthy of sustained attention.

ABSTRACT. This paper argues that the concept of intelligence is highly value-laden in ways that impact on the field of AI and debates about its risks and opportunities. This value-ladenness stems from the historical use of the concept of intelligence in the legitimation of dominance hierarchies. The paper first provides a brief overview of the history of this usage, looking at the role of intelligence in patriarchy, the logic of colonialism and scientific racism. It then highlights five ways in which this ideological legacy might be interacting with debates about AI and its risks and opportunities: 1) how some aspects of the AI debate perpetuate the fetishization of intelligence; 2) how the fetishization of intelligence impacts on diversity in the technology industry; 3) how certain hopes for AI perpetuate notions of technology and the mastery of nature; 4) how the association of intelligence with the professional class misdirects concerns about AI; and 5) how the equation of intelligence and dominance fosters fears of superintelligence. This paper therefore takes a first step in bringing together the literature on intelligence testing, eugenics and colonialism from a range of disciplines with that on the ethics and societal impact of AI.

ABSTRACT. One way of carving up the broad `AI ethics and society' research space that has emerged in recent years is to distinguish between ‘near-term’ and ‘long-term’ research. While such ways of breaking down the research space can be useful, we are concerned about the near/long-term distinction gaining too much prominence in how research questions and priorities are framed. We highlight some ambiguities and inconsistencies in how the distinction is used, and argue that while there are differing priorities within this broad research community, these differences are not well-captured by the near/long-term distinction. We unpack the near/long-term distinction into four different dimensions, and propose some ways that researchers can communicate more clearly about their work and priorities using these dimensions. We suggest that moving towards a more nuanced conversation about research priorities can help establish new opportunities for collaboration, aid the development of more consistent and coherent research agendas, and enable identification of previously neglected research areas.

ABSTRACT. We tackle the problem of building explainable recommendation systems that are based on a per-user decision tree, with decision rules that are based on single attribute values. We build the trees by applying learned regression functions to obtain the decision rules as well as the values at the leaf nodes. The regression functions receive as input the embedding of the user’s training set, as well as the embedding of the samples that arrive at the current node. The embedding and the regressors are learned end-to-end with a loss that encourages the decision rules to be sparse. By applying our method, we obtain a collaborative filtering solution that provides a direct explanation to every rating it provides. With regards to accuracy, it is competitive with other algorithms. However, as expected, explainability comes at a cost and the accuracy is typically slightly lower than the state of the art result reported in the literature. Our code is anonymously available at https://drive.google.com/file/d/1Yfs_cpAT3OKadaYo_5p25PaxFKvkFaP1/view?usp=sharing.

ABSTRACT. Given AI's growing role in modeling and improving decision-making, how and when to present users with feedback is an urgent topic to address. We empirically examined the effect of feedback from false AI on moral decision-making about donor kidney allocation. We found some evidence that judgments about whether a patient should receive a kidney can be influenced by feedback about participants' own decision-making perceived to be given by AI, even if the feedback is entirely random. We also discovered different effects between assessments presented as being from human experts and assessments presented as being from AI.

ABSTRACT. Fair decision-making in criminal justice relies on the recognition and incorporation of infinite shades of grey. In this paper, we detail how algorithmic risk assessment tools are counteractive to fair legal proceedings in social institutions where desired states of the world are contested ethically and practically. We provide a normative framework for assessing fair judicial decision-making, one that does not seek the elimination of human bias from decision-making as algorithmic fairness efforts currently focus on, but instead centers on sophisticating the incorporation of individualized or discretionary bias--a process that is requisitely human. Through analysis of a case study on social disadvantage, we use this framework to provide an assessment of potential features of consideration, such as political disempowerment and demographic exclusion, that are irreconcilable by current algorithmic efforts and recommend their incorporation in future reform.

ABSTRACT. Machine learning models are prone to biased decisions due to biases in the datasets they are trained on. In this paper, we introduce a novel data augmentation technique to create a fairer dataset for model training that could also lend itself to understanding the type of bias existing in the dataset i.e. if bias arises just from a lack of representation for a particular group (sampling bias) or if it arises because of human bias reflected in the labels (prejudice based bias). Given a dataset involving a protected attribute with a privileged and unprivileged group, we create an "ideal world" dataset: for every data sample, we create a new sample having the same features and label as the original sample but with the opposite protected attribute value. The synthetic data points are sorted in order of their proximity to the original training distribution and added successively to the real dataset to create intermediate datasets. We theoretically show that two different notions of fairness: statistical parity difference (independence) and average odds difference (separation) always change in the same direction using such an augmentation. We also show submodularity of the proposed fairness-aware augmentation approach that enables an efficient greedy algorithm. We empirically study the effect of training models on the intermediate datasets and show that this technique reduces the two bias measures while keeping the accuracy nearly constant for three datasets. We then discuss the implications of this study on the disambiguation of sample bias and prejudice based bias and discuss how pre-processing techniques should be evaluated in general. The proposed method can be used by policy makers who want to use unbiased datasets to train machine learning models for their applications to add a subset of synthetic points to an extent that they are comfortable with to mitigate unwanted bias.

ABSTRACT. To reduce human error and prejudice, many high-stakes decisions have been turned over to machine algorithms. However, recent research suggests that this \emph{does not} remove discrimination, and can perpetuate harmful stereotypes. While algorithms have been developed to improve fairness, they typically face at least one of three shortcomings: they are not interpretable, their prediction quality deteriorates quickly compared to unbiased equivalents, and %the methodology cannot easily extend other algorithms they are not easily transferable across models% (e.g., methods to reduce bias in random forests cannot be extended to neural networks) . To address these shortcomings, we propose a geometric method that removes correlations between data and any number of protected variables. Further, we can control the strength of debiasing through an adjustable parameter to address the trade-off between prediction quality and fairness. The resulting features are interpretable and can be used with many popular models, such as linear regression, random forest, and multilayer perceptrons. The resulting predictions are found to be more accurate and fair compared to several state-of-the-art fair AI algorithms across a variety of benchmark datasets. Our work shows that debiasing data is a simple and effective solution toward improving fairness.

ABSTRACT. Today, AI is increasingly being used in many high-stakes decision-making applications in which fairness is an important concern. Already, there are many examples of AI being biased and making questionable and unfair decisions. The AI research community has proposed many methods to measure and mitigate unwanted biases, but few of them involve inputs from human policy makers. We argue that because different fairness criteria sometimes cannot be simultaneously satisfied, and because achieving fairness often requires sacrificing other objectives such as model accuracy, it is key to acquire and adhere to human policy makers' preferences on how to make the tradeoff among these objectives. In this paper, we propose a framework and some exemplar methods for eliciting such preferences and for optimizing an AI model according to these preferences.

ABSTRACT. We propose a framework for analyzing and evaluating system-wide algorithmic fairness. The core idea is to use simulation techniques in order to extend the scope of current fairness assessments by incorporating context and feedback to a phenomenon of interest. By doing so, we expect to better understand the interaction among the social behavior giving rise to discrimination, automated decision making tools, and fairness-inspired statistical constraints. In particular, we invite the community to use agent based models as an explanatory tool for causal mechanisms of population level properties. We also propose embedding these into a reinforcement learning algorithm to find optimal actions for meaningful change. As an incentive for taking a system-wide approach , we show through a simple model of predictive policing and trials that if we limit our attention to one portion of the system, we may determine some blatantly unfair practices as fair, and be blind to overall unfairness.

ABSTRACT. As machine intelligence is increasingly incorporated into educational technologies, it becomes imperative for instructors and students to understand the potential flaws of the algorithms on which their systems rely. This paper describes the design and implementation of an interactive post-hoc explanation of the Bayesian Knowledge Tracing algorithm which is implemented in learning analytics systems used across the United States. After a user-centered design process to smooth out interaction design difficulties, we ran a controlled experiment to evaluate whether the interactive or `static' version of the explainable led to increased learning. Our results reveals that learning about an algorithm through an explainable depends on users' educational background. For other contexts, designers of post-hoc explainables must consider their users' educational background to best determine how to empower more informed decision-making with AI-enhanced systems.

ABSTRACT. Computer scientists have made great strides in characterizing different measures of algorithmic fairness, and showing that certain measures of fairness cannot be jointly satisfied. In this paper, I argue that the two most popular families of measures – target-conditional and score-conditional independence are actually best thought of as measures of the injustice of the contexts in which they are deployed. I begin by introducing three different ways of measuring bias - independence, prediction-conditional error and target-conditional error - and dis-cuss the implicit idealizations these measures make about the underlying causal structure of the contexts in which they are deployed. I then discuss three ways in which these idealizations fall apart in the con-text of deployment in an unjust world. In the final section, I suggest an alternative framework for measuring fairness in the context of existing injustice: justice-sensitive independence.

ABSTRACT. Work in Counterfactual Explanations tends to focus on the principle of “the closest possible world” that identifies small changes leading to the desired outcome. In this paper we argue that while this approach might initially seem intuitively appealing it exhibits shortcomings not addressed in the current literature. First, a counterfactual example generated by the state-of-the-art systems is not necessarily representative of the underlying data distribution, and may therefore prescribe unachievable goals (e.g., an unsuccessful life insurance applicant with severe disability may be advised to do more sports).Secondly, the counterfactuals may not be based on a “feasible path” between the current state of the subject and the suggested one, making actionable recourse infeasible (e.g., low-skilled unsuccessful mortgage applicants may be told to double their salary, which may be hard without first increasing their skill level). These two shortcomings may render counterfactual explanations impractical. To address these we propose a new line of Counterfactual Explanations research aimed at providing actionable and feasible paths to transform a selected instance into one that meets a certain goal. We address this challenge with FACE: an algorithmically sound way of uncovering these“feasible paths” based on the shortest path distances defined via density-weighted metrics. Our approach generates counterfactuals that are coherent with the underlying data distribution and supported by the “feasible paths” of change, which are achievable and can be tailored to the problem at hand.

ABSTRACT. In many judicial systems -- including the United States courts of appeals, the European Court of Justice, the UK Supreme Court and the Supreme Court of Canada -- for each case, a subset of judges is selected from the entire judicial body in order to hear the arguments and decide the judgment. Ideally, the subset selected is \emph{representative}, i.e., the decision of the subset would match what the decision of the entire judicial body would have been had they all weighed in on the case. Further, the process should be \emph{fair} in that all judges should have similar workloads, and the selection process should not allow for certain judge's opinions to be silenced or amplified via case assignments. Lastly, in order to be practical and trustworthy, the process should also be \emph{interpretable}, easy to use, and (if algorithmic) computationally \emph{efficient}. In this paper, we propose an algorithmic method for the judicial subset selection problem that satisfies all of the above criteria. The method satisfies fairness by design, and we prove that it has optimal representativeness asymptotically for a large range of parameters and under noisy information models about judge opinions -- something no existing methods can provably achieve. We then assess the benefits of our approach empirically by counterfactually comparing against the current practice and recent alternative algorithmic approaches using cases from the United States courts of appeals Database.

ABSTRACT. Numerous AI-assisted resource allocation decisions need to balance the conflicting goals of fairness and efficiency. Our paper studies the challenging task of defining and modeling a proper fairness-efficiency trade off. We define fairness with Rawlsian leximax fairness, which views the lexicographic maximum among all feasible outcomes as the most equitable; and define efficiency with Utilitarianism, which seeks to maximize the sum of utilities received by entities regardless of individual differences. Motivated by a justice-driven trade off principle: prioritize fairness to benefit the less advantaged unless too much efficiency is sacrificed, we propose a sequential optimization procedure to balance leximax fairness and utilitarianism in decision-making. Each iteration of our approach maximizes a social welfare function, and we provide a practical mixed integer/linear programming (MILP) formulation for each maximization problem. We illustrate our method on a budget allocation example. Compared with existing approaches of balancing equity and efficiency, our method is more interpretable in terms of parameter selection, and incorporates a strong equity criterion with a thoroughly balanced perspective.