Work in progress: no guarantees of accuracy, comprehensiveness, or style. Send a correction.

Canadian AI Safety Map

← All organizations

Aether AI Safety Research

Research · Toronto, ON, Canada

Control & oversightInterpretabilityAlignment

Independent technical AI safety group. Own words: "Aether is an AI safety research group. Our primary focus is on studying LLM monitoring and hidden reasoning, but we are also interested in understanding and shaping the propensities of LLMs." Work to date is chain-of-thought monitoring (how information access affects a monitor's ability to detect sabotage, ICML 2026), a taxonomy of hidden reasoning in LLMs, and a no-CoT task-completion time-horizon benchmark. Current focus is a "science of held-out monitors": when held-out measures of alignment silently degrade under training against correlated proxies, and when they remain trustworthy.

Website
https://aether-ai-research.org
People
Rauno Arike (Managing Director & Research Scientist), Shubhorup Biswas, Dennis Akar, Vladimir Ivanov, Joey Yudelson (Research Scientists), advisors: Rohan Subramani (founder; now at Coefficient Giving), Seth Herd (Astera Institute), Zhijing Jin (University of Toronto), Francis Rhys Ward (Arrow Research)
Founded
Not recorded
Source
https://aether-ai-research.org/assets/index-BZfejIrb.js
Confidence
AI verified Checked against the organization's own website on 2026-08-31.

Something wrong? Email us.