Researcher · Founder · Executive
Toby Shevlane
ManticCo-founder and CEO
Shevlane links AI safety research with forecasting. He co-authored Model Evaluation for Extreme Risks, on testing dangerous AI capabilities,[1] and developed “structured access”: letting people use AI while restricting unsafe uses.[2] At Mantic, he builds systems that forecast real-world events to improve decision-making.[3]
Google DeepMind
GovAI
University of Oxford
University of CambridgeCitation Trend
References
- Model evaluation for extreme risksarxiv.org
He co-authored Model Evaluation for Extreme Risks, on testing dangerous AI capabilities,
- Structured access: an emerging paradigm for safe AI deploymentarxiv.org
Sole-authored chapter develops the structured-access framework; does not imply all controlled-access practices originated with him.
- Mantic — official websitemantic.com
At Mantic, he builds systems that forecast real-world events to improve decision-making.
Based on public sources. If accuracy is important, verify details using the links or references.
For corrections, updates, or requests to remove information, contact us.
Reviewed November 9, 2026 · Checked for accuracy and new information.