What to study for each EBAI domain
An ethical framework for using AI
Covers the principles that form an ethical framework for AI, including fairness, transparency, accountability, explainability, and respect for human autonomy. It also addresses beneficence, non-maleficence, inclusivity, and how to balance competing principles and build justified trust.
Key skills
fairness, transparency, accountability, explainability, human oversight, autonomy, beneficence, non-maleficence, stakeholder identification, justified trust
Innovating ethically with AI to drive business change
Covers using AI to drive business change in an ethical way, from identifying genuine problems and assessing ethical risk to piloting responsibly and scaling on evidence. It also addresses impacts on staff and customers, leadership accountability, and measuring success beyond profit.
Key skills
ethical business case, pilot design, stakeholder consultation, staff impact and training, leadership accountability, success measures, responsible scaling, harm response
Data privacy, governance and policy in AI
Covers data privacy, governance, and policy for AI, including GDPR-style principles such as lawful basis, purpose limitation, and data minimisation. It also addresses individual rights, consent, privacy notices, impact assessments, retention, and breach response.
Key skills
lawful basis, purpose limitation, data minimisation, data rights, consent, privacy notices, impact assessments, retention, pseudonymisation, breach response
Data architecture, sustainability and ethics
Covers how data architecture choices affect ethics, including data quality, representativeness, provenance, lineage, access control, and dataset linkage risks. It also addresses the environmental impact of AI, including energy use, carbon measurement, efficient models, and sustainable procurement.
Key skills
data quality, representative data, provenance and lineage, access control, dataset linkage, carbon footprint, efficient models, sustainable procurement, retention and disposal
Ethically building and testing AI solutions
Covers ethically building and testing AI solutions, including bias testing, validation, robustness testing, and fairness evaluation across groups. It also addresses human-in-the-loop oversight, independent review, acceptance criteria, safe fallbacks, and post-deployment monitoring.
Key skills
bias testing, validation, robustness testing, fairness evaluation, human-in-the-loop, independent review, acceptance criteria, safe fallbacks, drift monitoring