No legal liability framework for AI misalignment harms, expert warns it's a disaster
LuizaJarovsky · x · 2026-11-16
AI governance researcher Luiza Jarovsky argues there is no clear way to allocate legal liability for harms caused by model misalignment, calling the gap a "disaster."
Key points:
- There is currently no technical fix for AI alignment; even with state-of-the-art security measures, misalignment incidents could still happen, and alignment may not even be possible.
- If jurisdictions hold AI labs legally liable for misalignment harms, the liability framework collides with the fact that the underlying technical problem remains unsolved.
The piece highlights the mismatch between legal regimes and the unresolved alignment problem.
More from AGI Musings
- Netflix lands Joseph Gordon-Levitt's AI thriller with Rachel McAdams, sparking debate on AI apocalypse blockbusters — ShakeelHashim · 2026-11-16
- Caine Prize confirms 2026 winner amid AI-use allegations — ShakeelHashim · 2026-11-16
- No legal liability framework exists for model misalignment harms, and that's a disaster — LuizaJarovsky · 2026-11-16
- A diagram for judging quality in cultural artefacts: coded meaning and exemplary moves — round · 2026-11-16
- Economists argue humans will be fine without work: jobs were just reputation and friendship engines — paulnovosad · 2026-11-16
- What Labs Actually Do With $100M Data Deals: Turning Company Workflows Into Agent Eval Sandboxes — StewartalsopIII · 2026-11-16