AI could break legal systems by finding previously hidden loopholes
Reported by The Register1 source ↓
Legal scholars say large language models could identify statutory inconsistencies at scale. They cite possible gaps in German law, but experts say human review is needed to confirm whether gaps are real or consequential.
Key points
- Ringe and Folkerts used large language models to identify potential gaps in German law
- Examples include a possible missing legal citation in an occupational-diseases rule and a transit visa rule
- Santa Clara law professor Eric Goldman says human review is needed to confirm AI-identified gaps
- The researchers shared only examples they believe could not cause serious harm
Details
Felix Ringe, a visiting fellow at Stanford Law School’s liftlab, and collaborator Linus Folkerts argue that large language models could find inconsistencies in written law that people might exploit. The Register reports that they used LLMs to identify possible gaps in German law, while cautioning that courts may interpret statutes to resolve textual problems.
One example concerns an amendment to Germany’s Occupational Diseases Regulation. The researchers argue it omits a constitutionally required citation to its statutory authority, which could potentially support a challenge to a confidentiality provision. They also describe a transit visa rule that may be invalid because the government did not identify the authority for making it.
Eric Goldman, a law professor at Santa Clara University, told The Register that using AI to identify possible statutory gaps is a logical approach, but people must determine whether the gaps are real and legally significant. He said legislatures may correct major gaps once identified; he also suggested legislative counsel could use AI to check draft laws.
Written from coverage as of
Images
- The Register
Stocks in this story
Change since open: 20 Nov to 21 Nov 2026
Last refresh