Web3 Has No Safe AI. DMind AI Just Quantified the Gap -- and KDD 2026 Made It Official.

Enhao Huang — Ph.D. Candidate, National Key Laboratory of Blockchain and Data Security, Zhejiang University; Lead Researcher, DMind Benchmark. Photo: DMind AI

The first peer-reviewed Web3 AI benchmark tests 31 top models — including GPT-5, Claude, and Gemini — across 3,543 expert questions. The verdict: no system is ready for the field's highest-stakes tasks.

SINGAPORE, May 31, 2026 /PRNewswire/ -- Medical AI has MedQA. Financial AI has FinBen. Legal AI has LegalBench. Web3 — one of the most adversarial, financially consequential software environments in existence — had nothing. Today, that changes.

Recommended for you