Connor Leahy 🇩🇪
AI safety researcher and ControlAI U.S. director who founded Conjecture and helped start EleutherAI, with work focused on advanced-AI control and governance.
ControlAI
- Evidence items
- 3
- Toward pressure
- +0.10
- Away pressure
- −0.03
- Net attributed pressure
- +0.07
0 comments · 0 votes
Sign in to join the discussion →
No comments yet. Start the discussion.
First-person and official sources
These sources guide discovery. A statement still needs a dated, attributable, source-backed evidence assessment before it can affect the index.
- institutional profileConjecture
- institutional profileConnor Leahy at ControlAI
Assessments involving Connor Leahy
Connor Leahy urges a verified international ban on superintelligence
In Canadian Senate testimony, Leahy argued that opaque frontier development and competitive pressure make uncontrolled superintelligence a national-security risk. He proposed a trust-but-verify international prohibition backed by domestic enforcement and sanctions, while opposing a broad slowdown of data-centre development.
Connor Leahy argues for an international AI red line in nuclear command
Connor Leahy argued that governments should prohibit dangerous AI uses in nuclear-war decisions and develop treaty-style monitoring of advanced AI. He linked the governance gap to systems receiving greater autonomy despite unresolved control and interpretability problems, while presenting nuclear non-proliferation as a model for international restraint.
Connor Leahy proposes deepfake liability as a first step toward AGI governance
In a dated TIME interview, Connor Leahy argued that liability across the deepfake development and distribution chain could build enforcement capacity for wider AI harms, alongside an internationally verifiable compute cap on frontier training runs.