Skip to content
Berkeley Existential Risk Initiative at a glanceMostly funded, not a grantmaker: see who funds it →

· Public charity

Berkeley Existential Risk Initiative

To improve human civilizations long-term prospects for survival and flourishing.

$2.5M
Granted FY2024still arriving
5
Grants FY2024still arriving
2
States reached
$1.5M
Largest
01What it fundsRead
02The needRead
03Its granteesRead
04Its fieldRead
01What it funds
01100% classified

What it funded, over time

Every grant placed by its stated purpose and the recipient’s mission, by year, FY2018 to 2024.

Science & Tech$2.6MPhilanthropy$2.5MHealth$1.6MEnvironment$1.0MEducation$746kSocial Science$50k
02FY2024 · 5 grants

Where the money goes

Grants by size, and the largest recipients.

How this is measured
The 5 grants below total $2,062,055, the rows itemized in this filing. The $2,541,655 headline is the total grant expense on the return; the other $479,600 is giving the schedule does not break out (grants under the $5,000 itemization floor, grants to individuals, grants on other schedules). Every figure below describes the itemized rows only.

By grant size · FY2024

  • $10k–50k2 grants · $41k
  • $250k+3 grants · $2.0M
$250,000
Median grant
2
States reached
$4.5M
Total assets
Largest grants
RecipientAmount
MATS RESEARCH INC$1,521,400
STANFORD UNIVERSITY$250,000
INTERNATIONAL ASSOCIATION FOR SAFE & ETHICAL AI$250,000
SECUREBIO INC$24,655
RETHINK PRIORITIES (FISCAL SPONSOR OF APOLLO RESEARCH)$16,000
02The need
03

Does the money reach high-need areas?

Each grant placed by the hardship in its grantee’s ZIP, against the area’s typical level.

show:

Its human-services grants (FY24–24, $16k) land where the poverty rate runs at 10%, against an area that typically sits at 9%. 100% of those dollars go to grantees based in above-average-need neighborhoods. Most of its grants land in higher-need ZIPs.

area typical 9%RETHINK PRIORITIES (FISCAL SPONSOR OF APOLLO RESEARCH): $16k → 10%0%20%40%50%more need →
grant to an above-average-need area below average· circle size = grant amount
How this is measured
Matched: human-services grants are those whose grantee’s IRS cause code is Human Services. Placed by people below the poverty line in the grantee’s ZIP, averaged from the tracts in that ZIP.

Grants abroad, by region: $286k on the FY2024 return

Schedule F, as filed: 5 regions, $220k to organizations and $66k to individuals. The IRS asks for region and purpose, not the recipient, so no country or grantee can be named here.

Middle East and North Africa$220k · 77% · 1 grant

Stated purpose: GRANT AI SAFETY SOLUTIONS

Europe (including Iceland and Greenland)$35k · 12% · 35 people

Stated purpose: PRIZES

East Asia and the Pacific$13k · 5% · 2 people

Stated purpose: PRIZES

Sub-Saharan Africa$9k · 3% · 20 people

Stated purpose: PRIZES

North America (Canada and Mexico)$9k · 3% · 5 people

Stated purpose: PRIZES

How this is measured
US grants placed by each grantee’s ZIP, which can differ from where services are delivered. Need from public data (U.S. Census/ACS, CDC PLACES, Eviction Lab, USDA), shown as context about the area, never attributed to the funder’s giving. Grants abroad on Schedule F are filed by region, purpose and amount with no recipient name, so they are shown by region and cannot be placed on the country map or matched to a grantee.
03Its grantees
repeat funding

Who it funds again

Organizations funded in more than one year, against those funded once.

48%of every dollar goes to organizations it has funded before.
$2.8M · 3 repeat orgs$3.0M to everyone else

Grantees it renews have grown median revenue +181% since the first grant, against +48% for those funded once.

3 repeat relationships: 0 still active in FY2024, 3 since wound down; 4 grantees were first funded in FY2024 (too recent to call).

How the two cohorts compare

Re-uppedFunded once

Organizations

3
10

Total granted

$2.8M
$1.2M

Median revenue growth · since first grant

+181%
+48%

Still filing today

100%
80%

New vs renewed · share of each year

In FY2024, 0% of grant dollars renewed an existing relationship; $1.8M went to new ones.

50%100%’18’19’20’21’22’23’24
RenewedFirst-time

Where new relationships form · theme of each grantee’s first grant

’18’19’20’21’22’23’24
Science & TechEducationHealthHuman ServicesEnvironmentSocial ScienceOther
How this is measured
First-time means a grantee’s first year in the filing window; renewed means funded in an earlier year too. The renewed share climbs naturally as a portfolio matures, so the signal is a year that bucks it (a wave of new grantees). The earliest year is left-censored: relationships that predate the data read as new. Theme is the grantee’s IRS cause.

Backed again, and grew

  • CF
    CENTER FOR APPLIED RATIONALITY
    3× · 2018–2022 · $1.7M · revenue +64% · 35% of their budget
  • SA
    SOCIAL AND ENVIRONMENTAL ENTREPRENEURS (SEE) INC
    4× · 2019–2022 · $1.0M · revenue +345%
  • EV
    EFFECTIVE VENTURES FOUNDATION USA INC
    2× · 2019–2023 · $56k · revenue +181%

Funded once

  • UO
    UNIVERSITY OF CALIFORNIA BERKELEY FOUNDATIONgraduated
    one grant, 2018 · $400k · revenue +99%
  • MI
    MACHINE INTELLIGENCE RESEARCH INSTITUTE INC
    one grant, 2020 · $300k · revenue -80%
  • FO
    FUTURE OF LIFE INSTITUTEgraduated
    one grant, 2018 · $300k · revenue +682%
How this is measured

Repeat means funded in two or more distinct years. Growth and survival are read from each grantee’s own subsequent IRS filings.

04Its field

Closest unfunded matches

The nonprofits that most resemble its grantees but are not funded by it, scored along the direction from the rest of the field toward the portfolio. A resemblance in what they say they do, not a recommendation.

1
Safe Artificial Intelligence Forum Institute

To drive global coordination that minimizes extreme AI risks and ensures the benefits of AI are shared by all. SAIF fosters responsible governance to reduce catastrophic risks through shared understanding and collaboration among key global…

—
2
Center for Ai Safety Inc

to reduce societal-scale risks associated with artificial intelligence AI by advancing safety research, building the field of AI safety researchers, and promoting safety standards.

Science & Tech
3
Clarity Ai Research Inc

Transluce is an independent research lab that builds open, scalable technology for understanding ai systems and steering them in the public interest.

Science & Tech
4
Polaris Research Institute Inc

The institute's primary purpose is to undertake and contribute to research to improve the quality of life of as many sentient beings as extensively as possible and to otherwise engage in activities in furtherance of this objective. As a…

Science & Tech
5
Iit Research Institute

Perform basic research and preclinical efficacy assessments of new molecular entities being developed for the prevention and treatment of cancer and infectious disease. Discover new pathways and targets for the design and development of…

Science & Tech
6
Palisade Research

Our mission is to research and communicate the capabilities and risks posed by advanced artificial intelligence systems. our current work focuses on cybersecurity, deception, and persuasion. we aim to communicate the state of these…

Science & Tech
7
The Eleutherai Institute

Our mission is to advance research on interpretability and alignment of foundation models and to educate about the capabilities, limitations, and risks associated with these technologies.

Science & Tech
8
Future House Inc

Our mission is to build an ai scientist, or ai systems, that can automate scientific research and accelerate the pace of discovery, so humanity can find cures for disease, solutions for climate change, and other species-accelerating…

Science & Tech
9
Future Tech Insights Inc

The organization provides public education on artificial intelligence.

Public Benefit
10
OpenAI Foundation fka OpenAI Inc

OpenAIs mission is to ensure that artificial general intelligence benefits all of humanity.

Science & Tech
11
Arc Research Institute

Our mission is to accelerate scientific progress, understand the root causes of disease, and narrow the gap between discoveries and impact on patients.

Health
12
Corporation for National Research Initiatives

Scientific research in information processing; fostering dev. of nat'l inf. infra.

Science & Tech
How this is measured
The grantee most central to the portfolio’s shape is Machine Intelligence Research Institute Inc; the most unlike its peers is Institute for Philosophical Research. Resemblance is measured on each organization’s own IRS 990 mission text; it reflects how work is described, not its quality or impact.
04the grantee network

18 grantees tracked through their own filings, 2017–2025.

Each grant beside the grantee’s revenue from every source, year by year.

How this is measured

Counted here: distinct organizations funded across 2017 to 2025, not grant rows in a single year, so this will not match the grant count on the cover. 18 of the 19 grantees resolved in that span have returns of their own we could reconcile.

Each grantee is resolved to its own IRS returns. Association, dated; never a causal claim.

2
Load-bearing (≥25% of a budget)
4
Early backer (in before they grew)
17/18
Grantees still filing
10/18
Grew since first funded

Where the money sits, by cause and then by grantee

Silicon Valley Community Foundation: $2,500,000 · PhilanthropySilicon Valley Community Foundati…MATS RESEARCH INC: $1,521,400 · Science & TechMATS RESEARCH INCMACHINE INTELLIGENCE RESEARCH INSTITUTE INC: $300,000 · Science & TechMACHINE INTELLIGENCE RESEARCH IN…FUTURE OF LIFE INSTITUTE: $300,000 · Science & TechFUTURE OF LIFE INSTITUTEINTERNATIONAL ASSOCIATION FOR SAFE AND ETHICAL ARTIFICIAL INTELLIGENCE INC: $250,000 · Science & TechINTERNATIONAL ASSOCIATION FOR SA…+4 more: $107,153 · Science & TechCENTER FOR APPLIED RATIONALITY: $1,720,125 · HealthCENTER FOR APPLIED RAT…SOCIAL AND ENVIRONMENTAL ENTREPRENEURS (SEE) INC: $1,022,223 · EnvironmentSOCIAL AND EN…UNIVERSITY OF CALIFORNIA BERKELEY FOUNDATION: $400,000 · EducationUNIVERSIT…THE BOARD OF TRUSTEES OF THE LELAND STANFORD JUNIOR UNIVERSITY: $250,000 · EducationTHE BOARD…EFFECTIVE VENTURES FOUNDATION USA INC: $56,316 · EducationEFFECTIVE…CONVERGENCE ANALYSIS: $39,282 · EducationINSTITUTE FOR PHILOSOPHICAL RESEARCH: $50,000 · OtherREGENTS OF THE UNIVERSITY OF CALIFORNIA AT BERKELEY: $26,000 · OtherLEVERAGE RESEARCH INC: $25,000 · Social ScienceRethink Priorities: $16,000 · Human Services
Philanthropy$2,500,000Science & Tech$2,478,553Health$1,720,125Environment$1,022,223Education$745,598Other$76,000Social Science$25,000Human Services$16,000

Each grantee by its size and the funder’s share of it; top left is load-bearing

25%50%75%100%$1.0M$10M$100Mgrantee revenue →↑ funder's share of their budgetSilicon Valley Community Foundation: $2,500,000 over 2y, 0.1% of budgetCENTER FOR APPLIED RATIONALITY: $1,720,125 over 3y, 35% of budgetMATS RESEARCH INC: $1,521,400 over 1y, 24% of budgetSOCIAL AND ENVIRONMENTAL ENTREPRENEURS (SEE) INC: $1,022,223 over 4y, 1.0% of budgetUNIVERSITY OF CALIFORNIA BERKELEY FOUNDATION: $400,000 over 1y, 0.1% of budgetMACHINE INTELLIGENCE RESEARCH INSTITUTE INC: $300,000 over 1y, 3.8% of budgetFUTURE OF LIFE INSTITUTE: $300,000 over 1y, 11% of budgetTHE BOARD OF TRUSTEES OF THE LELAND STANFORD JUNIOR UNIVERSITY: $250,000 over 1y, 0.0% of budgetEFFECTIVE VENTURES FOUNDATION USA INC: $56,316 over 2y, 0.2% of budgetINSTITUTE FOR PHILOSOPHICAL RESEARCH: $50,000 over 1y, 91% of budgetCONVERGENCE ANALYSIS: $39,282 over 1y, 23% of budgetAssociation for the Advancement of Artificial Inte: $25,000 over 1y, 0.8% of budgetLEVERAGE RESEARCH INC: $25,000 over 1y, 9.4% of budgetSECUREBIO INC: $24,655 over 1y, 0.3% of budgetRethink Priorities: $16,000 over 1y, 0.1% of budgetQuantified Uncertainty Research Inst: $7,498 over 1y, 2.6% of budget
Grantee by grantee: a decade per organization, before and after funding

A decade per grantee: revenue shaded, grants as bars, all rows on one timeline.

Co-funders

Grantmakers whose grantees overlap with this funder’s far more than size alone predicts. Closer means stronger alignment; arcs show where they also fund each other. Here they form a tightly interlocked group: most of these funders back each other's grantees too.

Founders Pledge IncSilicon Valley Commun…Good Ventures Foundat…Rethink CharityBeri Support FundEffective Ventures Fo…Vanguard Charitable E…Network for GoodPaypal Charitable Giv…National Philanthropi…American Online Givin…Fidelity Investments …
Founders Pledge IncNY26.3× affinity10 shared granteesties to 11 of 11Hover any node to trace its alignments.Compare side by side →

Open a dossier: Founders Pledge Inc · Silicon Valley Community Foundation · Good Ventures Foundation · Rethink Charity · Beri Support Fund · Effective Ventures Foundation USA Inc · Vanguard Charitable Endowment Program · Network for Good · Paypal Charitable Giving Fund · National Philanthropic Trust · American Online Giving Foundation Inc · Fidelity Investments Charitable Gift Fund

How this is measured
Affinity is a Gamma-Poisson posterior co-funding rate, re-centered on the typical rate, so thin evidence shrinks toward no signal. Overlap is association, not proof of shared intent.

How much its grantees rely on government

Each dot is one organization Berkeley Existential Risk Initiative funds: left to right is the share of its income from government, bottom to top the share from Berkeley Existential Risk Initiative. Bigger dots raise more.

2024
’20’21’22’23’24
no gov · 20%2%6%14%25%funder's share of their income ↑0%25%50%75%100%share of the org’s income from government →
    no gov moneyreceives it· size = income
    2get no government money at all
    1report government grants on their 990 we could not trace to a source (not plotted)
    0rely on government for over half their income
    typical government reliance, FY20–24
    How this is measured
    Government income is each org’s traced federal awards (USASpending grants and contracts) plus state payments (open checkbooks) as a share of its total revenue (IRS Form 990); the self-reported government-grant line (990 line 1e) is carried for cross-check: a grantee that reports government grants we could not trace to a source is left off the chart rather than shown as receiving none. An association, not a claim that a grant caused the public funding. Coverage is precision-first, a floor rather than a census; state records exist for 9 states, so a grantee outside them shows no state figure (dimmed) rather than a false zero. Federal award amounts are obligations, which can span years, so a single-year share is indicative. 990 filings lag 12 to 24 months.

    On method. Every financial figure here is read from IRS e-file XML, the funder’s own 990/990-PF, and from the multi-year returns of the 19 grantees we resolved, each linked to its source. Grantee outcomes are each organization’s own program-service reporting (Form 990, Part III), read as association with funding, never as caused by it. Low-confidence name matches are suppressed rather than guessed. How we build these →

    How this is measured
    Each figure carries one of six labels. Filed is what the funder reported on its return. Official is another government record about an organization, such as a federal award or a charity register, joined by name where no shared identifier exists. Resolved is an identity we worked out where the filing named a recipient without an EIN, kept only above a measured confidence threshold. Computed is arithmetic over those, like themes, portfolio clusters and co-funder strength. Context is a statistic about a place rather than about an organization, which is what the need overlay is: it describes the area a grantee’s address sits in, not where its work lands. Inferred is drawn by a model from text, like the partnerships read out of public news and organization websites. Each is labeled where it appears.

    Generated from the IRS Form 990 e-file return for fiscal year 2024, released 2024. Filings run roughly 12 to 24 months behind; figures are dated accordingly.

    Source object · view filing

    Play a game based on Berkeley Existential Risk Initiative’s filings →

    More from the funding graph

    The unnamed billions
    grants reported with no recipient in the public record
    The regranting layer
    51% of grant dollars land in the regranting layer
    The DAF merry-go-round
    $9.7B circulates between donor-advised funds
    The endowment market
    who holds foundation money, and what just moved
    The camps
    funders cluster by geography, not ideology
    By state
    where grant money is based, and what stays home