Intelligence Briefing
Co-founded DeepMind in 2010 with Demis Hassabis and Mustafa Suleyman. Popularized the term "artificial general intelligence" (AGI) with Ben Goertzel in the early 2000s. Leads DeepMind's Technical AGI Safety team. His PhD thesis under Marcus Hutter on machine super intelligence (AIXI) was foundational to AGI theory. Recently stated that the time for AGI skepticism is over, warning that AI will eliminate many remote cognitive jobs as systems surpass human-level capabilities.
BS, Computer Science — University of Waikato
MS, Computer Science — University of Auckland
PhD, Machine Super Intelligence — IDSIA / Università della Svizzera italiana
Operational History
Continued AGI Research
Continues to lead research on AGI safety and development at DeepMind.
careerWarning on Job Displacement
Legg warned that AI advancements would eliminate many remote cognitive jobs.
policyPhD Thesis Publication
Published his PhD thesis on machine super intelligence, contributing to AGI theory.
researchAGI Safety Advocacy
Legg emphasized the urgency of AGI safety in various public forums.
policyAlphaGo's Success
DeepMind's AlphaGo defeated the world champion Go player, showcasing advanced AI capabilities.
researchPublication of AIXI
Published foundational work on AIXI, a theoretical framework for AGI.
researchDeepMind Acquisition by Google
DeepMind was acquired by Google, marking a significant milestone in AI research.
careerCo-founding of DeepMind
Shane Legg co-founded DeepMind with Demis Hassabis and Mustafa Suleyman.
foundingAGI Position Assessment
2025-2030
Strong advocate for AGI safety and preparedness.
- AGI is approaching rapidly.
- Safety measures must be prioritized.
- Existential risks from AGI are significant.
Proactive and comprehensive AGI safety research.
Intercepted Communications
“The time for AGI skepticism is over; we must prepare for the implications of advanced AI.”
“AI will eliminate many remote cognitive jobs as systems surpass human-level capabilities.”
“AGI safety is not just a priority; it is an urgent necessity.”
“The foundational work on AIXI is crucial for understanding machine intelligence.”
“We are on the brink of a new era in AI; the implications are profound.”
Research Output
Advancements in Machine Intelligence
2023AI Research Review
Recent advancements in machine intelligence.
The Future of AGI Safety
2022AI Safety Conference
Discussed the importance of AGI safety.
Ethics in AI Development
2021AI Ethics Journal
Addressed ethical considerations in AI.
Machine Super Intelligence
2020PhD Thesis
PhD thesis that laid groundwork for AGI theory.
Reinforcement Learning and AGI
2019Journal of AI Research
Explored the relationship between reinforcement learning and AGI.
AIXI: A Universal Agent
2015Journal of Machine Learning Research
Foundational paper on AGI.
DeepMind's Research Agenda
2010DeepMind Publications
Foundational research agenda for DeepMind.
Universal Intelligence: A Definition
2007Artificial Intelligence
Introduced the concept of universal intelligence.
Known Associates
Demis Hassabis
co-founderCo-founded DeepMind together.
View Dossier →Mustafa Suleyman
co-founderCo-founded DeepMind and worked on AGI safety.
View Dossier →Ben Goertzel
collaboratorCollaborated on the concept of universal intelligence.
View Dossier →Marcus Hutter
mentorSupervised Legg's PhD thesis on machine super intelligence.
View Dossier →Organizational Affiliations
Current
Google DeepMind
Co-founder & Chief AGI Scientist
2010 - Present
Former
IDSIA
Researcher
2008 - 2010
University of Lugano
Postdoctoral Researcher
2007 - 2008
Source Material
Dossier last updated: 2026-03-04