Skip to main content
See every side of every news story
Published loading...Updated

Waterloo Researchers Use Math to Verify AI Safety

University of Waterloo researchers developed AI tools that mathematically verify safety in critical systems like power grids, matching or exceeding traditional methods, Liu said.

  • On October 21, 2025, University of Waterloo researchers published in Automatica a study on math-based verification of AI controllers, using DOI 112193, to address AI safety concerns.
  • As AI moves into critical infrastructure, validation efforts often get short shrift due to a fast-deployment mindset, leaving safety gaps in power grids and autonomous vehicles.
  • Liu's group trained neural networks to learn safety proofs, using Lyapunov functions to predict stability, while a logic-based reasoning system verifies these proofs rigorously.
  • Tests show the framework matched or exceeded traditional methods, and Liu's team is developing an open-source toolbox while exploring industry collaborators within Waterloo's TRuST Scholarly Network and federal responsible-AI efforts.
  • It may sound surprising that AI checks AI, but Dr. Jun Liu said, `To be clear, no one is attempting to create factories or systems run entirely by AI without any human input`, and these tools ensure safety guarantees for physical systems.
Insights by Ground AI

13 Articles

Think freely.Subscribe and get full access to Ground NewsSubscriptions start at $9.99/yearSubscribe

Bias Distribution

  • 100% of the sources are Center
100% Center

Factuality 

To view factuality data please Upgrade to Premium

Ownership

To view ownership data please Upgrade to Vantage

City News broke the news in Toronto, Canada on Thursday, October 2, 2025.
Sources are mostly out of (0)
News
For You
Search
BlindspotLocal