Category: TheRAIN Recommended Read
-
Deputy Prime Minister speech on AI for Public Good
Deputy Prime Minister Oliver Dowden recently outlined a vision for the UK to become a global leader in applying Artificial Intelligence (AI) to public services. In a speech at Imperial College London, Dowden highlighted the potential of AI to revolutionize everything from healthcare to education, promising faster services, reduced costs,…
-
Operationalising AI governance through ethics‑based auditing:
AI ethics audit struggles: New research finds practical challenges beyond existing frameworks. Multinational companies face internal obstacles like aligning standards, defining scope, and driving change. Study suggests EBA needs more than just evaluation metrics to be effective. Read article here >
-
The Ethical Guidelines for Trustworthy AI – A Procrastination of Effective Law Enforcement
EU AI ethics guidelines reviewed: diversity concerns, “ethics shopping” risks, and enforcement gaps identified. Red lines urged, citing social/ecological costs, research buyouts, and neglected ethical issues. Law enforcement focus seen as essential next step for responsible AI development. Read article here >
-
Towards AI ethics’ institutionalization: knowledge bridges from business ethics to advance organizational AI ethics
AI ethics need lessons from business ethics! This paper proposes applying knowledge from areas like stakeholder management and corporate governance to address AI’s ethical challenges. By sharing insights, both fields can learn and improve, preventing “ethics washing” and promoting responsible AI development. Read article here >
-
Anthropomorphism in AI: hype and fallacy
Assigning human qualities to AI (anthropomorphism) overstates its abilities and muddies our moral judgments, leading to ethical issues. This essay explores how attributing human-like traits can both exaggerate AI’s performance and distort our understanding of its moral responsibilities. Read article here >
-
Adopting trust as an ex post approach to privacy
AI accuracy crucial for privacy trust: research delves into how individuals and AI systems can collaborate to ensure information sharing stays private. Trustworthiness requires shared beliefs, and an AI’s accuracy impacts individual competence and thus, trustworthiness. This affects privacy directly, highlighting the need for regulations that consider the relationship between…
-
To be forgotten or to be fair: unveiling fairness implications of machine unlearning methods
Machine unlearning, a tool for RTBF (Right to be Forgotten), can impact AI fairness. This study explores two common methods (SISA, AmnesiacML) vs. retraining (ORTR) across fairness datasets and deletion strategies. Results show non-uniform deletion with SISA yields better fairness outcomes, while other methods have mixed effects. These findings inform…
-
Seeing Like a Toolkit: How Toolkits Envision the Work of AI Ethics
AI ethics toolkits promise guidance, but fall short! They lack support for navigating complex power dynamics within organizations, leaving practitioners unprepared to tackle real-world ethical challenges. Future toolkits need to address these issues for truly responsible AI development. Read article here >
-
From ethical AI frameworks to tools: a review of approaches
AI ethics: Drowning in principles, lacking solutions. Many guidelines exist, but they’re too vague and don’t translate well to real-world AI. This analysis highlights the need for concrete methods and tools that address specific ethical issues beyond just “explainability, fairness, privacy, and accountability.” We need to move from principles to…
-
The uselessness of AI ethics
AI ethics are drowning in a sea of ineffective principles. While guidelines abound, they’re often vague, ignored, and lack teeth, failing to address the real dangers of AI. We need better solutions! Instead of relying solely on principles, we must tackle systemic issues of oppression and prioritize rigorous accuracy checks.…
RAIN 