Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
April 15, 2026AI & SocietyOpen Access

Justifiability and AI: putting explainability in its place

View Full Paper
Ask AI
Bookmark
Share

Authors

BBBoris BabicICI. Glenn CohenJSJulian Savulescu

Discussion

Loading...

Member takes

Overview

Explores AI/ML justifiability in sensitive decision-making contexts, indicating its importance over mere explainability.

Key Points

  • The study aims to differentiate between AI/ML explainability and justifiability, advocating for the latter as a valuable objective.
  • Analysis of explainability and justifiability concepts in AI/ML systems
  • Comparison between motivating and normative reasons
  • Exploration of various criteria for algorithmic justifiability
  • Justifiable models can be trusted even when they lack full explainability
  • Post hoc explainability algorithms offer limited value compared to justifiability
  • Effective justifications rely solely on normative reasons, unlike explanations which require motivating reasons

Cite This Study

Babic et al. (2026) studied this question.

synapsesocial.com/papers/69df2b2ce4eeef8a2a6b0282https://doi.org/10.1007/s00146-026-03030-9
View Full Paper
Ask AI
Bookmark
Share