Update README.md

README.md

@@ -184,7 +184,6 @@ Full results are reported in the accompanying paper; key findings:
 ## Limitations
 
 - **GDPR categories are not legal advice.** G-pillar flags indicate potential regulatory risk, not a legal determination of non-compliance. Human review is required for enforcement decisions.
-- **GDPR violation detection is hard.** Recall on a dedicated GDPR violations dataset remains low (~13–31%) even with grounding, reflecting that G1/G2 violations often require contextual legal reasoning beyond surface text features. This is consistent with broader findings in legal NLP.
 - **Synthetic-to-real gap.** In-domain F1 is inflated by shared generation artifacts. External benchmarks provide a more conservative estimate of real-world performance.
 - **European Portuguese focus.** The GDPR categories were developed and evaluated for pt-PT. Coverage of edge cases in other EU languages is untested.