Forecasting Research Institute
Forecasting Research Institute is a nonprofit research organization that runs large-scale expert forecasting studies on AI, biosecurity, and existential risk, distributing findings freely to policymakers, researchers, and the public through its website, Substack, and academic publications.
- Company typePrivate
- Founded2013
- Headquarters—
- Headcount1–10
- GTM typeB2B
- OfferingServices
What Forecasting Research Institute does
Forecasting Research Institute (FRI) is a Delaware-incorporated nonprofit research organization that advances the science of forecasting to inform decision-making on high-stakes issues including AI trajectory, biosecurity, nuclear risk, and existential risk. The institute runs large-scale expert elicitation studies and develops novel forecasting methodologies, most notably the Longitudinal Expert AI Panel (LEAP) — a three-year monthly survey tracking 339 leading experts across computer science, AI industry, economics, and policy — and ForecastBench, a dynamic benchmark measuring AI forecasting accuracy published at ICLR 2025. Its methodology portfolio also includes the Reciprocal Scoring method for forecasting unanswerable questions and Conditional Trees for question generation, both developed under Chief Scientist Philip Tetlock, who pioneered first-generation forecasting research.
FRI's core technology stack is centered on structured survey-based prediction, judgmental forecasting, and expert elicitation, processing over 600,000 words of expert rationales across the first three LEAP waves alone. Products and findings are distributed freely through the institute's website, the LEAP portal, ForecastBench, a Substack newsletter, academic preprints on SSRN and arXiv, and earned media coverage in outlets including The New York Times, The Economist, Bloomberg, TIME, Vox, and Marginal Revolution.
The institute is funded primarily through philanthropic grants, with Open Philanthropy as the primary funder and Founder's Pledge members providing additional support for specific projects. FRI does not sell commercial products or services; its outputs are freely available to policymakers, academic researchers, AI industry leaders, and the public. The organization co-leads LEAP in partnership with Stanford University, the Federal Reserve Bank of Chicago, and the University of Pennsylvania, with additional support from Princeton AI Lab.
Forecasting Research Institute firmographics
Firmographics- Name
- Forecasting Research Institute
- Legal name
- Forecasting Research Institute
- Website
- https://forecastingresearch.org
- Company type
- Private
- Founded year
- 2013
- Operating status
- Operating
- Headcount range
- 1–10 employees
- Short description
- Forecasting Research Institute is a nonprofit research organization that runs large-scale expert forecasting studies on AI, biosecurity, and existential risk, distributing findings freely to policymakers, researchers, and the public through its website, Substack, and academic publications.
- Ownership category
- akta.pro rank
Forecasting Research Institute industry classification
Industry- Product category
- Forecasting Research and Expert Elicitation
- NAICS
- Research and Development in the Social Sciences and Humanities (54172), Research and Development in the Social Sciences and Humanities (541720), Other Scientific and Technical Consulting Services (54169)
- SIC
- Services-Commercial Physical & Biological Research (8731)
- akta.pro primary industry
- Scenario Planning, Forecasting Platforms & Planning Intelligence (HDAAAHAL)
- akta.pro secondary industries
- AI Governance, Risk & Compliance (GRC) Platforms (HDAAAMAA), Model Monitoring, Accountability & Incident Management (HDAAAMAF), Model Monitoring, Drift & Behavioral Anomaly Detection (HDAAAKAF)
Keywords
Forecasting Research Institute business model
Business model- GTM type
- B2B
- Offering type
- Services
- Cost components
- Personnel, Technology or R&D, Operations, Marketing or Sales
Revenue model
- Grant Funding and Philanthropic Support: FRI is a nonprofit organization incorporated in Delaware. The institute's research is funded primarily by philanthropic organizations, with Open Philanthropy explicitly cited as a key supporter. Additional support acknowledged includes Founder's Pledge members for specific research projects. Revenue is used to fund research operations, forecasting studies, and researcher compensation.
Go-to-market motion1 record
Distribution channels3 records
Marketing channels6 records
Forecasting Research Institute product offering
Product offeringCore offering
Forecasting Research Institute designs and operates large-scale expert forecasting studies on high-stakes issues including AI, biosecurity, and nuclear risk. It runs the Longitudinal Expert AI Panel (LEAP), the ForecastBench AI accuracy benchmark, the Reciprocal Scoring and Conditional Trees methodologies, and specialized research initiatives on existential and catastrophic risks, publishing findings as free reports via its website, Substack, and academic repositories.
Product overview
Forecasting Research Institute (FRI) operates a research platform focused on advancing the science of forecasting for high-stakes decision-making. The core product is the Longitudinal Expert AI Panel (LEAP), a three-year monthly survey tracking expert predictions on AI's trajectory across science, economy, and society. Related offerings include ForecastBench, a dynamic benchmark measuring AI forecasting accuracy; the Reciprocal Scoring method for forecasting unanswerable questions; and specialized research initiatives on existential risk, biorisk, and nuclear risk. The portfolio comprises primary research products (LEAP, ForecastBench), methodological tools (Reciprocal Scoring), and domain-specific studies (biorisk forecasting, nuclear risk assessment, existential risk projects).
Differentiator
Problem solved
Functional benefit
Brands
- LEAP (Longitudinal Expert AI Panel): A three-year project tracking expert predictions on the effects of artificial intelligence on society, conducted in partnership with Stanford University, Federal Reserve Bank of Chicago, and University of Pennsylvania.
- ForecastBench
Products and services
- Longitudinal Expert AI Panel (LEAP) A three-year monthly survey project tracking views of leading computer scientists, industry professionals, policy researchers, and economists on the trajectory of artificial intelligence; participants provide thousands of forecasts on AI progress indicators, labor market impacts, and scientific discovery.
- ForecastBench A dynamic, continuously updated benchmark measuring the forecasting accuracy of AI systems on a changing set of questions, remaining contamination-free over time and benchmarked against human superforecasters; the associated paper was published at ICLR 2025.
- Reciprocal Scoring Method An elicitation method that challenges forecasters to predict the forecasts of other forecasters, creating accountability for accuracy in policy debates on otherwise unanswerable questions; two studies demonstrated RS forecasters were as accurate as those predicting objectively resolvable outcomes using proper scoring rules.
- Nuclear Risk Forecasting Study The largest systematic survey of subject matter experts on nuclear weapons risk, predicting the likelihood of nuclear conflict and the impact of specific tractable policies on nuclear catastrophe probability; published as a working paper in October 2024.
- Forecasting LLM-enabled Biorisk and the Efficacy of Safeguards A forecasting study examining expert views on AI-enabled biosecurity threats, assessing how advancing LLM capabilities might increase risks of human-caused epidemics and evaluating the efficacy of mitigation measures.
- Improving Judgments of Existential Risk Research improving alignment between forecasting tournaments and actionable forecasts of existential risk by measuring facets of human judgment through intersubjective metrics such as Reciprocal Scoring.
Quantifiable outcome
- Experts predict 18% of U.S. work hours will be assisted by generative AI by 2030, up from ~4.1% in November 2024 — a 4x increase
- +5 more outcomes
Companies that use Forecasting Research Institute
Customer profileNamed customers3 records
Segments4 records
Ideal customer profiles3 records
Forecasting Research Institute technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- No
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability5 records
Feature4 records
Forecasting Research Institute partnerships and signals
Strategic signalPartnerships
Four partnerships are on record, tiered core and supporting.
- Stanford UniversitycoreStanford University researchers co-lead the Longitudinal Expert AI Panel (LEAP) project alongside FRI. Stanford's involvement brings academic credibility and expertise to the monthly expert forecasting surveys tracking AI trajectory and societal impact.
- Federal Reserve Bank of ChicagocoreResearchers at the Federal Reserve Bank of Chicago co-lead the LEAP project. Ezra Karger serves as both a researcher at FRI and at the Federal Reserve Bank of Chicago, contributing economic expertise and forecasting methodology to the initiative.
- University of PennsylvaniacoreUniversity of Pennsylvania researchers co-lead LEAP alongside FRI, Stanford, and the Federal Reserve Bank of Chicago. Philip E. Tetlock, Chief Scientist at FRI, holds a position at the Wharton School of the University of Pennsylvania.
- Princeton AI LabsupportingPrinceton AI Lab provides support to the LEAP project, contributing to the panel's research on AI trajectory and expert forecasting.
Scale indicators5 records
Recent moves6 records
Expansion highlights6 records
Forecasting Research Institute competitors and assessment
Company assessmentMarket position
Competitive moat5 records
Key risks5 records
Key highlights7 records
Customer concentration
Forecasting Research Institute social profiles
Digital presenceForecasting Research Institute financial estimates
Financial estimateRevenue estimate
Valuation estimate
Forecasting Research Institute leadership team
Management profileNumber of profiles
Profiles10 records
Forecasting Research Institute funding detail
Funding detailFunding overview
Funding rounds
Investors
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Forecasting Research Institute M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Forecasting Research Institute
What does Forecasting Research Institute do?
Forecasting Research Institute designs and operates large-scale expert forecasting studies on high-stakes issues including AI, biosecurity, and nuclear risk. It runs the Longitudinal Expert AI Panel (LEAP), the ForecastBench AI accuracy benchmark, the Reciprocal Scoring and Conditional Trees methodologies, and specialized research initiatives on existential and catastrophic risks, publishing findings as free reports via its website, Substack, and academic repositories.
Is Forecasting Research Institute a public or private company?
Forecasting Research Institute is a private company. It is classified as nonprofit foundation owned and is currently operating.
When was Forecasting Research Institute founded?
Forecasting Research Institute was founded in 2013. It employs 1 to 10 people.
How does Forecasting Research Institute make money?
One revenue line is on record: grant Funding and Philanthropic Support.
Does Forecasting Research Institute have an API?
No public API is recorded for Forecasting Research Institute.
What industry is Forecasting Research Institute in?
Forecasting Research Institute's product category is Forecasting Research and Expert Elicitation. Its primary akta.pro industry code is HDAAAHAL, Scenario Planning, Forecasting Platforms & Planning Intelligence, with a secondary code of HDAAAMAA, AI Governance, Risk & Compliance (GRC) Platforms. Its NAICS code is 54172 and its SIC code is 8731.