Floridi et al. paper: intelligence, black box, hallucination metaphors skew AI law
ArtificialOther · x · 2026-11-09
A new paper by Claudio Novelli, Luciano Floridi, Stefan Larsson, Mariarosaria Taddeo and Steven Winter in the International Journal for the Semiotics of Law uses cognitive linguistics to show how metaphor shapes AI regulation.
- Law doesn't meet AI as a blank slate; it reaches for imaginative patterns that frame risks, allocate authority, and create doctrinal path dependence — not mere rhetoric.
- The authors dissect three influential concepts: "(artificial) intelligence" invites personification and misplaced assumptions about agency (what's produced is a simulation of narrow linguistic competence); "black box" misleadingly turns a distributed sociotechnical system into a bounded container; and "hallucination" steers regulatory attention in unhelpful ways.
More from Safety
- Tencent AI chief Yao Shunyu warns AI will soon master biology, unleashing biosecurity risks — pstAsiatech · 2026-11-09
- Running vulnerability scanners on global legislation to surface legal loopholes — SnooPeripherals5313 · 2026-11-09
- Infosec researcher maps 3 low-hanging abuse paths for open-weight models — matthew_d_green · 2026-11-09
- AI-assisted attacks are noisy, not smart: same recon command 400 times, says researcher — matthew_d_green · 2026-11-09
- Transluce builds pipeline uncovering thousands of innocuous-looking datasets that induce harmful behavior after finetuning — RexDouglass · 2026-11-09
- Proof launches agent identity layer with x401, letting AI agents prove who authorized them — Scobleizer · 2026-11-09