Ethics Guidelines for Trustworthy AI
European Commission
The 2019 guidelines from the European Commission's High-Level Expert Group on AI setting out seven requirements for trustworthy AI, including human oversight, privacy, transparency, fairness, and accountability, plus an assessment list for applying them to real AI systems.
More resources on Big Data & Ethics
Bit by Bit: Social Research in the Digital Age — Chapter 6: Ethics
Free online chapter from a Princeton sociologist's textbook on research with large digital datasets. Works through three contested studies, four principles from the Belmont and Menlo reports, informed consent, informational risk and privacy, so readers can reason through ethically ambiguous data projects. Fills the gap the other picks leave: a principles-based decision framework for data collection and use (consent, informational risk, re-identification) rather than model fairness.
Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy
Non-technical book by a former hedge-fund quant examining how opaque scoring models in credit, hiring, policing, insurance and education harm people at scale. Readers come away able to recognise the traits of a harmful model: opacity, scale and damaging feedback loops. The standard entry-point book on big-data harms and a direct critical counterweight to the catalog's Mayer-Schönberger & Cukier 'Big Data'.
Fairness and Machine Learning: Limitations and Opportunities
The field's reference text on algorithmic bias: graduate-level textbook, free online in full, on fairness in automated decision-making. Covers legitimacy of automated decisions, formal fairness criteria, causality, US anti-discrimination law, testing for discrimination, and datasets. Readers can formally evaluate a model for bias and understand each criterion's limits.
Practical Data Ethics
Six-lesson video course with syllabus and reading list, originally taught at the University of San Francisco Data Institute. Covers disinformation, bias and fairness, ethical foundations, privacy and surveillance, and industry incentives, giving learners tools to spot and mitigate data misuse. The one free, complete, no-prerequisite course that spans the whole topic note (bias, privacy, governance incentives) with case studies and a curated reading list.
Big Data
An eleven-module Coursera course from O.P. Jindal Global University covering Hadoop, HDFS, MapReduce, Hive, Pig, HBase and Apache Spark. Graded assignments use PySpark, Spark SQL and MLlib on real datasets, so you finish able to build distributed processing pipelines.
Big Data
Viktor Mayer-Schönberger and Kenneth Cukier explain how massive datasets shift analysis from sampling and causation toward correlation, with examples from business, health, and government. Readers come away understanding big data's practical promise and its risks for privacy, prediction-based punishment, and accountability.