European Union Regulations on Algorithmic Decision Making and a “Right to Explanation”
2016/06/28 by Bryce Goodman, Seth Flaxman · 10 voices · 90 citations
Social Sciences · Computer Science · Medicine · #Ethics and Social Impacts of AI #Law, AI, and Intellectual Property #Artificial Intelligence in Healthcare and Education
paper · pdf · doi:10.1609/aimag.v38i3.2741
Abstract
We summarize the potential impact that the European Union's new General Data Protection Regulation will have on the routine use of machine‐learning algorithms. Slated to take effect as law across the European Union in 2018, it will place restrictions on automated individual decision making (that is, algorithms that make decisions based on user‐level predictors) that “significantly affect” users. When put into practice, the law may also effectively create a right to explanation, whereby a user can ask for an explanation of an algorithmic decision that significantly affects them. We argue that while this law may pose large challenges for industry, it highlights opportunities for computer scientists to take the lead in designing algorithms and evaluation frameworks that avoid discrimination and enable explanation.
Cited by
- Counterfactuals and Causability in Explainable Artificial Intelligence: Theory, Algorithms, and Applications
- The Mythos of Model Interpretability
- Relation between prognostics predictor evaluation metrics and local interpretability SHAP values
- Balancing explainability and privacy in AI systems: A strategic imperative for managers
- Toward Human-Understandable, Explainable AI
- A Unifying Review of Deep and Shallow Anomaly Detection
- Transparency in Complex Computational Systems
- Artificial intelligence in orthopaedics surgery: transforming technological innovation in patient care and surgical training
- COVAriance-Induced Fairness Gap Penalty for Subgroup-Fair Clustering
- Footprints of Data in a Classifier: Understanding the Privacy Risks and Solution Strategies
- Machine Learning Testing: Survey, Landscapes and Horizons
- GANMEX: One-vs-One Attributions Guided by GAN-based Counterfactual Explanation Baselines
- Computing Optimal Decision Sets with SAT
- Extracting spatial effects from machine learning model using local interpretation method: An example of SHAP and XGBoost
- Information Theoretic Measures for Fairness-aware Feature Selection
- Explainable Automated Coding of Clinical Notes using Hierarchical Label-wise Attention Networks and Label Embedding Initialisation
- Concept-based Visual Counterfactual Explanations with Diffusion Models
- Scalable Explanation of Inferences on Large Graphs
- Towards Fair Affective Robotics: Continual Learning for Mitigating Bias in Facial Expression and Action Unit Recognition
- Imputation Uncertainty in Interpretable Machine Learning Methods
- A Framework for Causal Concept-based Model Explanations
- CryptoSPN: Privacy-preserving Sum-Product Network Inference
- GDPR Anti-Patterns: How Design and Operation of Modern Cloud-scale\n Systems Conflict with GDPR
- Bias Amplification in Artificial Intelligence Systems
- The effectiveness of feature attribution methods and its correlation\n with automatic evaluation scores
- Interpretable Deep Convolutional Fuzzy Classifier
- SecureBoost: A Lossless Federated Learning Framework
- The effects of explainability and causability on perception, trust, and acceptance: Implications for explainable AI
- Characterizing Intersectional Group Fairness with Worst-Case Comparisons
- State of the Art in Fair ML: From Moral Philosophy and Legislation to\n Fair Classifiers
- Evaluating neural network explanation methods using hybrid documents and morphological agreement
- Trustworthy AI: From Principles to Practices
- Scientific Exploration and Explainable Artificial Intelligence
- The role of artificial intelligence for management decision: a structured literature review
- The Need for Standardized Explainability
- A Sparsity Algorithm with Applications to Corporate Credit Rating
- Discrimination, intelligence artificielle et decisions algorithmiques
- Discrimination, artificial intelligence, and algorithmic decision-making
- Exact and Consistent Interpretation of Piecewise Linear Models Hidden behind APIs: A Closed Form Solution
- From black box to clarity: Strategies for effective AI informed consent in healthcare
- The Answer Lies Within: Self-Derived Rewards Enable Explainable Relation Extraction
- Expl(AI)ned: The Impact of Explainable Artificial Intelligence on Users’ Information Processing
- Causal Interpretability for Machine Learning -- Problems, Methods and Evaluation
- Structural Refinement of Bayesian Networks for Efficient Model Parameterisation
- A systematic review of algorithm aversion in augmented decision making
- Black-box artificial intelligence: an epistemological and critical analysis
- CausalKANs: interpretable treatment effect estimation with Kolmogorov-Arnold networks
- Learning Conformal Explainers for Image Classifiers
- Adaptive Cybersecurity Architecture for Digital Product Ecosystems Using Agentic AI
- Training Feedforward Neural Networks with Standard Logistic Activations\n is Feasible
- Better together? Human oversight as means to achieve fairness in the European AI Act governance
- Interpretability of neural networks via polynomial representations
- Posthoc Interpretability of Learning to Rank Models using Secondary Training Data
- Exact and Consistent Interpretation for Piecewise Linear Neural Networks: A Closed Form Solution
- BMB-LIME: LIME with modeling local nonlinearity and uncertainty in explainability
- Truthful AI: Developing and governing AI that does not lie
- Transparent and Fair Profiling in Employment Services: Evidence from Switzerland
- Progressive Disclosure: Designing for Effective Transparency
- Minimal Data, Maximum Clarity: A Heuristic for Explaining Optimization
- Trustworthy Artificial Intelligence: A Review
- Explainable AI and Adoption of Financial Algorithmic Advisors: an\n Experimental Study
- Evaluating the stability of model explanations in instance-dependent cost-sensitive credit scoring
- Deep opacity and AI: A threat to XAI and to privacy protection mechanisms
- Extending LIME for Business Process Automation
- Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learning
- Detecting Statistical Interactions from Neural Network Weights
- AI and Agile Software Development: A Research Roadmap from the XP2025 Workshop
- Transparency, Auditability and eXplainability of Machine Learning Models\n in Credit Scoring
- Explainable AI (XAI): A systematic meta-survey of current challenges and future opportunities
- An interpretable semi-supervised classifier using two different\n strategies for amended self-labeling
- A Novel Study on Intelligent Methods and Explainable AI for Dynamic Malware Analysis
- Towards Interpretable Deep Extreme Multi-label Learning
- Collective dynamics of strategic classification
- Iterative Learning of Computable Phenotypes for Treatment Resistant Hypertension using Large Language Models
- Fairwashing: the risk of rationalization
- Interpreting Classifiers through Attribute Interactions in Datasets
- Combinatorial Approaches for Embedded Feature Selection in Nonlinear SVMs
- Explaining Natural Language Processing Classifiers with Occlusion and Language Modeling
- Self-learn to Explain Siamese Networks Robustly
- A copula-based visualization technique for a neural network
- An Overview on the Landscape of R Packages for Credit Scoring
- Outlining the Design Space of Explainable Intelligent Systems for Medical Diagnosis
- Understanding Black-box Predictions via Influence Functions
- Machine Learning Testing: Survey, Landscapes and Horizons
- Ethical Considerations in Artificial Intelligence Courses
- Metrics for Explainable AI: Challenges and Prospects
- Incorporating Priors with Feature Attribution on Text Classification
- Beyond Fairness Metrics: Roadblocks and Challenges for Ethical AI in Practice
- Towards A Rigorous Science of Interpretable Machine Learning
- Artificial intelligence [wikipedia]
Discussions
- EU regulations on algorithmic decision-making and a “right to explanation” [hn, 104 points, 120 comments]
- EU regulations on algorithmic decision-making and a "right to explanation" [lobsters, 21 points, 11 comments]
- EU regulations on algorithmic decision-making and a “right to explanation” [hn, 4 points, 0 comments]
- EU regulations on algorithmic decision-making and a “right to explanation” [hn, 2 points, 0 comments]
- EU regulations on algorithmic decision-making and a “right to explanation” [hn, 2 points, 0 comments]
- EU regulations on algorithmic decision-making and a “right to explanation” [hn, 2 points, 0 comments]
- EU regulations on algorithmic decision-making and a “right to explanation” [hn, 1 points, 0 comments]
- EU regulations on algorithmic decision-making and a “right to explanation” [hn, 1 points, 0 comments]
- EU law to go into effect in 2018 will establish "right to explanation" for machine learning based decisions. PDF: https://arxiv.org/pdf/1606.08813v1.pdf [bsky, 0 points, 0 comments]
- "EU regulations on algorithmic decision-making and a 'right to explanation'", by Bryce Goodman and Seth Flaxman. http://arxiv.org/pdf/1606.08813v1.pdf [bsky, 0 points, 0 comments]
Related