Center on Long-Term Risk

What Center on Long-Term Risk does

The Center on Long-Term Risk (CLR) is a UK-registered research charity (England and Wales, no. 1195079) founded in July 2013 (originally as the Foundational Research Institute, renamed March 2020) that addresses what it terms "s-risks" — risks of astronomical suffering — from advanced AI systems, with a focus on two threat models: cooperation failures between AI systems (catastrophic bargaining failures, AI flash wars, commitment races) and harmful propensities in AI systems (s-risk conducive properties, or SRCPs). Its outputs comprise formal research (open-source game theory, evolutionary game theory, decision-theoretic frameworks), empirical LLM persona studies, and macrostrategy analysis. The organization runs the Model Persona Research Agenda and Safe Pareto Improvements Research Agenda, maintains a publications library and blog, and operates the CLR Fund for grantmaking in adjacent work. CLR is staffed by approximately 12 core personnel and advisors, led by Managing Director Tristan Cook; it is headquartered in London with a satellite footprint across the effective altruism and AI safety ecosystems (Oxford, Bay Area, Berlin historically).

CLR's revenue model is purely donation-based: funding flows from individual donors and a small set of institutional funders — Open Philanthropy, Macroscopic Ventures, Survival and Flourishing Fund, and Community Foundation Ireland — with tax-deductible giving enabled via Giving What We Can (US/NL) and Effektiv Spenden (DE/CHE). Notably, a 2020 Open Philanthropy grant of $1M over two years to parent entity EAF underwrote early operations; subsequent budgets have been in the high-hundreds-of-thousands to ~$1M range, with a 2025 fundraiser targeting $400K for 2026. The go-to-market motion is community-led: CLR builds pipeline through workshops, the Summer Research Fellowship, the Research Affiliates Program, the SPI Fundamentals Program, and career coaching, with content distribution via blog, newsletter (~70 researchers), and annual review/fundraiser campaigns. Its primary customers/audience are AI safety researchers and academics working on alignment, cooperative AI, and long-term risk; the organization also serves individual donors and talent (career-coaching clients, fellowship applicants) within the EA and AI safety communities.

Name

Center on Long-Term Risk

Legal name

Center on Long-Term Risk

Website

https://longtermrisk.org

Company type

Private

Founded year

2013

Operating status

Operating

Headcount range

11–50 employees

Short description

Center on Long-Term Risk is a London-based nonprofit research charity that conducts interdisciplinary research and grantmaking to reduce s-risks (risks of astronomical suffering) from advanced AI, serving AI safety researchers, the effective altruism community, and individual donors via its research programs, CLR Fund, and fellowship programs.

Ownership category

akta.pro rank

Product category

AI Safety Research and Grantmaking

NAICS

Research and Development in the Social Sciences and Humanities (541720)

SIC

Services-Commercial Physical & Biological Research (8731)

akta.pro primary industry

Safety & Alignment Evaluation (red-teaming, harmful capability testing) (HDAAAMAL)

akta.pro secondary industries

LLM/GenAI Guardrails & Safety Controls (policy, filtering, routing) (HDAAAKAD), Responsible AI, Security & Privacy Platforms (Safety, Guardrails, PII) (HDAEANAG), Documentation & Reporting Automation (Model Cards, Datasheets, Impact Assessments) (HDAAAMAJ)

Keywords

• AI safety research
• S-risk reduction
• Grantmaking programs
• Research fellowships
• AI cooperation analysis

Headquarters

HQ city

London

HQ country

United Kingdom

HQ region

Europe

Offices2 records

Markets served

GTM type

B2C

Offering type

Services

Cost components

Personnel, Operations, Marketing or Sales, Others

Revenue model

1. Donations and grants: CLR is funded entirely by donations from individuals and institutional donors. No commercial revenue streams. Major institutional funders include Macroscopic Ventures, Open Philanthropy, Survival and Flourishing Fund, and Community Foundation Ireland. UK donors can claim Gift Aid. US and Netherlands donors can give tax-deductibly via Giving What We Can. Germany and Switzerland donors can give via Effektiv Spenden.

Go-to-market motion1 record

Distribution channels3 records

Marketing channels6 records

Core offering

Center on Long-Term Risk conducts interdisciplinary research, makes and recommends grants, and builds a community of researchers focused on reducing s-risks (risks of astronomical suffering) from advanced AI. Its work spans two primary threat areas: harmful propensities in AI systems (s-risk conducive properties) and catastrophic cooperation failures between AI systems (e.g., AI bargaining failures, flash wars). Outputs include research agendas, peer-reviewed and workshop papers, a grantmaking fund, fellowships, and educational programs.

Product overview

The Center on Long-Term Risk is a research organization rather than a product company. Its offerings consist of research programs (Model Persona Research Agenda and Safe Pareto Improvements Research Agenda), community building initiatives (Summer Research Fellowship, Research Affiliates Program, SPI Fundamentals Program), grantmaking through the CLR Fund, advisory services (Career Advice), and knowledge products (Publications and Blog). These offerings collectively form a portfolio of research, educational, and community-building activities aimed at reducing s-risks from advanced AI systems.

Differentiator

Problem solved

Functional benefit

Products and services

• Model Persona Research Agenda A research program focused on understanding and steering the emergence of s-risk conducive properties (SRCPs) in AI systems, examining how future powerful AIs might create or perpetuate suffering. Targets AI safety researchers and academics.
• Safe Pareto Improvements Research Agenda A research agenda examining strategies to prevent catastrophic cooperation failures between powerful AI systems, including credible commitments and bargaining mechanisms. Targets AI safety researchers and academics.
• CLR Fund A grantmaking initiative that supports work relevant to reducing s-risks, including funding researchers and projects aligned with CLR's mission. For researchers, nonprofits, and projects aligned with s-risk reduction.
• Summer Research Fellowship An annual fellowship program hosting students and researchers to work on s-risk reduction and cooperative AI research under CLR mentorship. Targets graduate students and early-career researchers.
• Research Affiliates Program A program connecting external researchers with CLR's research priorities and community of professionals working on long-term AI risks. Targets external AI safety researchers.
• Safe Pareto Improvements (SPI) Fundamentals Program A fundamentals program designed to introduce participants to concepts related to safe Pareto improvements and reducing cooperation failures in AI contexts. Targets individuals entering the AI safety field.
• Career Advice Career coaching and advice services for individuals interested in pursuing careers focused on s-risk reduction and AI safety research. Targets prospective and early-career AI safety professionals.
• Publications CLR's published research reports, working papers, and academic publications on topics including cooperative AI, decision theory, and s-risk macrostrategy. Targets AI safety researchers and the broader research community.
• Blog CLR's blog featuring articles on AI bargaining, transformative AI misuse risks, s-risks, and research updates from the organization. Targets the AI safety and effective altruism communities.

Quantifiable outcome

• 12 published research reports in 2021; paper accepted at two NeurIPS workshops
• +2 more outcomes

Segments1 record

Ideal customer profiles2 records

Technology focussed No

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

Partnerships

Four partnerships are on record, tiered core, notable and minor.

• Effective Altruism Foundation (EAF)coreStrategic or Co-development PartnerCLR was initially founded as a project of EAF. Operations were transferred to an independent UK charity (CLR) at the end of 2021. EAF ceased operations in 2025. EAF was the parent organization and provided initial operational infrastructure, including office space and administrative support.
• Giving What We CannotableChannel Partner/ Reseller/ DistributorGiving What We Can serves as a donation platform enabling US and Netherlands donors to give tax-deductibly to CLR. As of October 2025, this applies to CLR only (not CLR Fund). The platform also accepts donations in cryptocurrency on CLR's behalf and stocks from the USA.
• Effektiv SpendenminorChannel Partner/ Reseller/ DistributorEffektiv Spenden enables German and Swiss donors to make tax-deductible donations to CLR for amounts over €20,000.
• Center for Emerging Risk Research (CERR)notableStrategic or Co-development PartnerIn 2021, CLR contributed to the scale-up of the field of Cooperative AI through advising CERR. This was part of CLR's grantmaking and advisory activities.

Scale indicators6 records

Recent moves8 records

Expansion highlights6 records

Broad incumbents

• Anthropic: Major AI lab with substantial investment in alignment and safety research. Broad incumbent conducting large-scale safety work; CLR's research on AI cooperation failures intersects with Anthropic's responsible scaling and Constitutional AI agendas.
• DeepMind (Safety team): Alphabet-owned AI research lab with a dedicated safety research division. Broad incumbent participating in CLR workshops (per public records); operates at vastly larger scale on overlapping themes of cooperation, multi-agent systems, and alignment.
• OpenAI (Superalignment / Safety team): Frontier AI developer with a dedicated safety research team. Broader incumbent that works on AI alignment, interpretability, and safety, including topics overlapping CLR's mandate. Vastly better funded but operating in adjacent problem space.

Direct peers

• Future of Life Institute (FLI): Nonprofit focused on reducing existential risks from advanced technologies, especially AI. Comparable as a mission-aligned research and advocacy organization working on long-term AI risk, though with a broader technology risk portfolio than CLR.
• Rethink Priorities: Research and field-building organization within the effective altruism ecosystem. Directly comparable to CLR as a small EA-aligned research nonprofit with grantmaking-adjacent activities; CLR's fractional COO previously served as COO at Rethink Priorities.
• Redwood Research: AI safety research nonprofit focused on alignment and deceptiveness. Comparable as a research-focused AI safety organization with a similar team size and funding profile, though focused on near-term alignment rather than s-risks.
• Conjecture: AI safety research and deployment company focused on alignment and catastrophe risk. Comparable as a small, mission-driven AI safety organization working on theoretical and applied alignment, though with a more applied/commercial bent than CLR.
• Machine Intelligence Research Institute (MIRI): One of the original AI safety research organizations, MIRI focuses on alignment and decision theory for advanced AI. Directly comparable to CLR as a small, research-focused nonprofit working on long-term AI risk, though with broader alignment scope rather than s-risks specifically.
• Centre for the Governance of AI (GovAI): Research organization focused on AI governance and policy. Directly comparable as a research-focused AI safety nonprofit; CLR's fractional COO Abraham Rowe is also COO at GovAI, evidencing tight ecosystem overlap. Both are London-based and operate in the AI safety talent network.
• Stanford Center for AI Safety: Academic research center focused on AI safety, alignment, and societal-scale risks. Comparable as a safety-focused research organization operating on related conceptual problems, though embedded in an academic institution rather than an independent nonprofit.

Market position

Strengths4 records

Weaknesses4 records

Competitive moat4 records

Key risks5 records

Key highlights6 records

Customer concentration

• Facebookfacebook.com/LongTermRisk
• LinkedInlinkedin.com/company/center-on-long-term-risk

Compliance2 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles4 records

Subsidiaries1 record

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Center on Long-Term Risk