This is a working draft, and funders and organizations aren’t monoliths: every placement here is a judgment call made from public grants, open calls and filings. If we have something wrong, or you think an assertion here is incorrect, please let us know at kate@higherground.institute. We look forward to feedback on what we’re getting wrong.
Why a second guide
The political guide sorts people by coalition. Funders sort by what they think AI will do to institutions and where they think a dollar changes the outcome. This guide reads grants and open calls as evidence of those beliefs.
What they believe
Is AI a risk to contain, a tool to deploy, or a shift in power?
Who they fund
Researchers, civil society, government, or builders.
Which way they push
Slow it down, steer it, or speed it up.
How to spot them
Vocabulary gives them away. The decoder below sorts it by camp.
The seven camps at a glance
Each camp lists the political faction it sits closest to, so the two guides connect. The two populist factions are folded together here because neither has a meaningful philanthropic base.
| Camp | Core belief | Key people and orgs | Key funding |
|---|---|---|---|
| CounterweightsNearest faction: AI Ethics, AI Populists | AI concentrates power in a few companies. Philanthropy should pay for checks on that power: accountability research, advocacy and the civil society groups that do both. | Heather Gerken (Ford), John Palfrey (MacArthur), Martin Tisné (Omidyar’s AI Collaborative); grantees AI Now, DAIR, Data & Society |
|
| Public buildersNearest faction: AI Ethics, Pragmatists | AI infrastructure should exist outside the companies: public compute, open models, public-interest data. Don’t just check the labs, build the alternative. | Mark Surman (Mozilla), Martin Tisné (Current AI), the Public AI Network; Empire AI in New York; the CalCompute consortium in California |
|
| State capacityNearest faction: Pragmatists | Government has to be able to use and govern AI itself, or democracy loses public trust. Fund the people and systems that let agencies deliver. This camp is about how government works, not how big it should be: it has no position on what the state should do, only on whether it can deliver. | Jennifer Pahlka (Recoding America), Tom Kalil and Kumar Garg (Renaissance Philanthropy); Packard, Rockefeller, Arnold Ventures |
|
| Security establishmentNearest faction: Pragmatists, Tech Right seam | AI is a national security issue and a contest with China. Fund the experts and staff who will run that policy. | Hewlett; Georgetown CSET; Horizon Institute (places fellows in government); Founders Pledge |
|
| Safety scienceNearest faction: AI Safety | Advanced AI could cause catastrophe. Fund research and researchers to prevent it, and grow the field quickly. | Dustin Moskovitz and Cari Tuna (Coefficient Giving), Jaan Tallinn (SFF), Jed McCaleb (Navigation Fund), Eric and Wendy Schmidt, Yoshua Bengio (LawZero) |
|
| Discovery acceleratorsNearest faction: Tech Right | AI will speed science and growth. Fund the breakthroughs and clear the barriers to building. | Chan Zuckerberg Initiative, XTX Markets, Stand Together (Koch network), Tyler Cowen (Emergent Ventures), Joe Lonsdale |
|
| Lab foundationsNearest faction: none yet | AI companies should give back part of their gains, mostly to help people adopt AI and to soften job losses. They stay out of policy and politics. | OpenAI Foundation (Bret Taylor, chair; Anna Makanju; Paul Christiano on the board); Anthropic Public Benefit Fund (leadership not yet named) |
|
Placements are preliminary. Fund managers such as Renaissance Philanthropy and pooled funds such as Humanity AI are treated as delivery vehicles; the camp follows the money behind them.
What the grants show
Four findings that don’t fit the two-sides story about AI money.
The same foundations fund both sides
Packard and Hewlett fund the accountability groups; Knight funds information and civic-capacity research; all three fund the center-right Foundation for American Innovation, which in turn helps steer AI safety grants.
Most new funds trace back to the same few foundations
Four of the five newer international funds traced back to the same funders: Omidyar, Ford, MacArthur, Hewlett and Open Society. Omidyar alone sits behind all four.
AI profits now pay for AI safety
The Navigation Fund is funded by the AI computing company it started, now merged with Lightning AI in a $2.5B deal.
Populist organizing is supported by safety-aligned groups
AI Populist money mostly moves through political committees such as Guardrails and Tomorrow Together, and MAGA Populists have no philanthropic base of their own. Much of the organizing aimed at populist audiences is supported by safety-aligned groups instead. The Pro-Human AI Coalition was co-founded by the Future of Life Institute, Frank McCourt’s Project Liberty and Ergo Impact.
Migration pathways
Grants from the past two years show three shifts. All three end up at the same question: who holds power as AI spreads.
Legacy foundations are moving toward national security
Hewlett co-founded the 2017 Ethics and Governance of AI Fund. In 2026 it launched a $100M security initiative.
Safety funders now ask who controls AI
Longview ran a 2026 call on extreme power concentration. SFF added new Freedom and Fairness tracks. An addition, not a swap: the catastrophic-risk concern hasn’t softened.
Democracy funders are betting on government that works
Packard funds government capacity alongside accountability. Rockefeller now co-funds with the OpenAI Foundation.
Where the three shifts meet
Packard’s field report points to “gradual disempowerment” (see the decoder) as common ground between safety funders and accountability funders. Both groups now pay for independent testing and corporate accountability.
Bridge funders
These funders make grants in more than one camp, and most of the migration above runs through them.
| Funder | Money lands in | What that looks like |
|---|---|---|
| Packard | CounterweightsState capacitySecurity | Humanity AI member; funds work to reimagine how the federal government operates, which will have to include AI; funds the Horizon fellowship. The “democracy that delivers” framing in its AI and Democracy report covers all three. |
| Hewlett | SecurityCounterweights | The security initiative at home; founding funder of the Global Fund for a New Economy and a European AI & Society Fund member abroad. |
| Omidyar | CounterweightsPublic builders | Funds through more vehicles than anyone else: its own AI Collaborative (Martin Tisné), Humanity AI, Current AI, the Global Fund for a New Economy, the European AI & Society Fund and Reset.Tech. |
| Coefficient Giving | Safety scienceSecurity | The largest AI safety funder; also seeded Georgetown CSET and funds the Institute for Progress and FAI. |
| Rockefeller | State capacityLab foundations | AI readiness work in all 50 states; co-funds Cities Deliver with the OpenAI Foundation. |
| Renaissance Philanthropy | State capacityDiscovery | Manages funds for other donors, including government innovation funds and the AI for Math Fund; over $530M mobilized, plus an AI Talent Accelerator that places engineers inside partner organizations. Also funds FAI and Horizon. |
| Knight | State capacityCounterweights | Funded the institute that published “AI as Normal Technology”; the Knight Research Network, its 2019 bet of about $50M on tech-and-democracy research that grew out of the disinformation fight and now covers AI; $3M+ for city AI fellows; also an FAI funder. |
Where the money lands
Funders don’t fund camps. They fund organizations, and sorting the grantees the same way shows which ones stay in one camp and which ones take money from several. The cross-camp grantees are where the migration above actually happens.
| Grantee | Camps funding it | Who pays | What it tells you |
|---|---|---|---|
| Foundation for American Innovation | State capacitySecuritySafety scienceDiscovery | Packard, Hewlett, Knight, Coefficient Giving, SFF, Renaissance Philanthropy, the Recoding America Fund, Arnold Ventures, Stand Together, plus Bradley, Scaife and Searle; a16z, Anthropic, OpenAI and Meta as sponsors | Funded across more camps than other grantees here. A center-right think tank that sits in the middle of the network, with supporters from most of the camps |
| Horizon Institute for Public Service | Safety scienceState capacityDiscovery | Coefficient, Longview ($2.1M in 2024 and 2025), Navigation Fund, Packard, Renaissance, the Lonsdale family fund | Fellows placed in government is the product every camp wants. The pipeline was built with safety money and now runs on state-capacity money too |
| Georgetown CSET | Safety scienceSecurity | Open Philanthropy, now Coefficient: more than $100M for 2019 to 2025; Hewlett; the Public Interest Technology University Network | The original safety-to-security bridge, built with one grant |
| Institute for Progress | Safety scienceState capacityDiscovery | Coefficient ($9M over three years from 2023, now through its $120M Abundance and Growth Fund), Arnold Ventures, Astera, Sloan, Heising-Simons, Smith Richardson | Abundance is where Safety money and Discovery money agree |
| Center for Humane Technology | CounterweightsSafety science | Omidyar, Ford, Open Society, Knight, Craig Newmark; Packard ($600K in 2025); Good Ventures ($750K in 2025 for AI safety communications) | The social media reform brand, now carrying Moskovitz money and a seat in the pro-human coalition. The clearest Counterweights-to-Safety crossover |
| Encode | Safety scienceCounterweights | Future of Life Institute ($250K in 2024), SFF, Heising-Simons, Omidyar and the Omidyar-led Responsible Technology Youth Power Fund; takes no corporate or frontier-lab executive money | A youth group with both camps’ money, and the first grantee to be targeted for it: OpenAI subpoenaed Encode over its funders in 2025 |
| Americans for Responsible Innovation | Safety science | Coefficient Giving | The DC shop that built the 2025 coalition against federal preemption of state AI laws. Where safety grants turn into lobbying |
| Public AI Network | Public buildersState capacity | Mozilla, Aspen Digital, Open Future, Rockefeller, Patrick J. McGovern Foundation, Bertelsmann | “Public AI” as infrastructure draws state-capacity and AI-for-good money that accountability work doesn’t. Also a pro-human coalition member |
| Humans First | Safety science | Started in 2026 with a loan from the Center for AI Safety, per NBC News; chair Amy Kremer says the donors are private individuals “who believe in AI safety.” No donor list | A MAGA-branded anti-AI group. Populist organizing with safety-world origins |
| Irreplaceable | Safety science | Launched August 2026 by Phil Aroneanu, a 350.org co-founder; runs an Education Fund and an Action Fund and discloses no donors; a strategist also serves at the Center for AI Safety, per The Atlantic | Climate-movement organizers with reported safety-world ties. Student walkouts at 100+ colleges in October |
| AI Now Institute | Counterweights | Open Society, Ford, Mozilla, Omidyar, Luminate; MacArthur and Minderoo earlier; dropped Microsoft Research and DeepMind | The control case. The accountability flagship takes no safety or lab-adjacent money, and its funder list is the Counterweights camp in miniature |
Funder lists come from each organization’s own supporters page, grant announcements and Candid records, and are not exhaustive. Camp tags follow the funder, not the grantee’s own politics.
Pooled funds
- Humanity AI: $500M over five years from ten foundations; one application, ten funders
- European AI & Society Fund: 18 funders, including Omidyar, Hewlett and Open Society
- Responsible Technology Youth Power Fund: Omidyar plus ten others, including McGovern and Pivotal; Encode’s funder
Fund managers
- Renaissance Philanthropy: runs other donors’ funds; shows up on FAI’s and Horizon’s supporter lists
- Recoding America Fund: a fund that is itself a funder, including of FAI
- Longview and SFF: move new-wealth safety money, Tallinn’s included, into grantees like Horizon and Encode
Grantees that regrant
- Future of Life Institute: realized roughly $500M from Vitalik Buterin’s 2021 crypto gift, by his account; now funds Encode, Common Sense Media and the Institute for Family Studies
- Omidyar’s AI Collaborative: grants still being traced; Omidyar took its grantee list off its site in 2025
The other pipe is social. New donors get sorted by the rooms they join: donor communities such as Forward Global, Solidaire and the Democracy Alliance, and Stanford’s First Principles Forum for new tech wealth. Roy Bahat’s advice to the AI-rich is to pick a theory of change before a vehicle, which is why some of this money will arrive as political giving, angel checks and new organizations rather than grants.
Where grants become politics
Five of the grantees above have crossed from research into campaigns. FLI ran an $8M state campaign in 2026 and co-founded the pro-human coalition. ARI helped build the coalition that beat federal preemption of state AI laws, 99 to 1 in the Senate. Encode worked state bills and got subpoenaed by OpenAI for it; Omidyar, named as its funder, stopped signing public letters and pulled its grantee list. The Center for AI Safety’s c4, the CAIS Action Fund, co-sponsored California’s SB 1047 and lobbied for the US AI Safety Institute. Coefficient has started a program to train AI safety political operatives. By Politico’s tally, safety-affiliated groups outspent the industry side $44.4M to $27.9M in the 2026 midterms. That is the seam between this guide and the political one: the same foundations that fund the research now sit one step behind the PACs.
The crossover has its own critics inside the money. Vitalik Buterin said in March 2026 that FLI made “cultural and political action” its primary method after his gift, and warned that big-money political campaigns invite backlash.
Decoder: how this world talks
Each camp has its own vocabulary, and most of it sounds more neutral than it is.
| Term | What it means | Who says it | What it signals |
|---|---|---|---|
| Transformative AI | AI that changes the economy and society on the scale of the industrial revolution | Coefficient Giving, safety funders | Safety lineage, even when the word “safety” never appears |
| Field-building | Funding the people, fellowships and new organizations that grow a discipline | Coefficient, SFF, Longview | A long time horizon; betting on people over projects |
| AGI-pilled | Convinced that very powerful AI is coming soon | Safety and lab circles, often about legacy foundations | The divide over timelines. Packard’s report quotes one interviewee wishing philanthropy were “more AGI-pilled” |
| Normal technology | AI will spread slowly and unevenly, the way electricity did | Narayanan and Kapoor; Princeton CITP; Knight Institute | Skeptical of both the hype and the doom; favors existing institutions |
| Public interest AI | AI built or governed for public benefit, outside the profit motive | Omidyar, Current AI, MacArthur, Ford | Counterweights when it means accountability; Public builders when it means public alternatives to the labs |
| People-centered AI | Humanity AI’s banner: put people’s agency and rights first | Humanity AI coalition | Legacy foundations focused on present-day harms |
| Pro-human | Keep humans in charge; no superintelligence until proven safe | Future of Life Institute, now the Pro-Human AI Coalition with Project Liberty and Omidyar | Safety goals with populist tactics; the banner that crossed camps in 2026 (see the pro- words) |
| State capacity | Government’s ability to actually deliver, including the people and the procurement rules to do it | Pahlka, Recoding America, Niskanen, IFP | State capacity and the abundance movement |
| Democracy that delivers | Democracies keep trust only by producing results people feel | Packard’s AI and Democracy report | Links democracy funding to government capacity |
| Political economy of AI | Who owns the chips, data and infrastructure, and who benefits | Global Fund for a New Economy, AI Now, Packard’s report | Counterweights moving into antitrust and labor |
| Gradual disempowerment | People slowly handing control to AI and the firms that own it | Safety researchers; adopted by Packard’s report | The bridge term between safety and accountability funders |
| Third wave / philanthropic VC | Startup-style giving by new AI wealth: big, fast bets, often on new organizations rather than existing grantees | Nan Ransohoff, Renaissance, Project Tailwind; pushback from Jeff Bradach, Lyel Resner, Nathan Chappell and Jennifer Pahlka | New money that prefers building to grantmaking. The critics’ reply: households with $500M+ give a flat 1.2% of assets a year, so the wave is pledged, not deployed. Semafor and the FT both note that effective altruism, written off after FTX, is the movement best placed to receive it |
| TradPhil | Shorthand for traditional philanthropy: legacy foundations seen as slow, process-heavy and risk-averse. Usually said with a sigh | New AI wealth and safety circles, about legacy foundations. Ransohoff’s “Belief 3,” that traditional philanthropy won’t be able to keep up, is the polite version | The same divide as “AGI-pilled” from the money side: new donors who think the old donors are too slow for the moment |
| Pledged vs. structural capital | A pledge is a promise; structural capital is money that has actually left the donor’s hands and can be planned around | Nick Martin; Nathan Chappell’s “promises are not capital” | The test to apply to every third-wave number. Only one living couple from the original Giving Pledge cohort has been identified as fulfilling it |
| The business is the philanthropy | The company, not a foundation, is the vehicle for solving public problems, so building it counts as giving | Associated with Musk and Bezos; Craig Shapiro’s “Philanthropy’s Other Third Wave”; the Giving Pledge defectors of 2026 | Discovery accelerators and the Tech Right. Why some new AI wealth will never show up in a grants database |
| Regrantor / pooled fund | An intermediary that moves other donors’ money | Manifund, Humanity AI, the European AI & Society Fund | The original donor is one step removed, so trace the money back |
“Safety”
- Safety science: preventing catastrophe
- Counterweights: harm to people now
- Security: national security
- Lab foundations: responsible rollout
“Capacity”
- State capacity: government that delivers
- Safety science: a pipeline of researchers
- Counterweights: stronger civil society groups
- Lab foundations: cities delivering services with AI (Cities Deliver)
“Public interest”
- Counterweights: accountable to the public
- Public builders: owned and run outside the companies
- Lab foundations: the company’s own giving
- State capacity: public services that work
The pro- words: whose side is “good AI” on?
Every camp now has a phrase that means “AI that is good for people.” The phrases are not interchangeable. Each one tells you who is talking, what they are afraid of, and where their money goes.
| Phrase | What it means | Who says it | What it signals |
|---|---|---|---|
| Pro-social AI | AI and platforms designed to strengthen trust, cooperation and social cohesion rather than engagement at any cost. Borrowed from behavioral science (“prosocial behavior”) | Omidyar (“pro-social outcomes” in its Responsible Tech stance); the social media reform world: Prosocial Design Network, Lisa Schirch’s Council on Technology and Social Cohesion, Berkeley CHAI’s Prosocial Ranking Challenge, Glen Weyl and Audrey Tang’s “prosocial media” | Counterweights, design wing. Still a researcher’s word, not a funder’s banner. No fund or coalition is named for it |
| Pro-human AI | Humans stay in charge: no superintelligence until proven safe, no concentration of power, liability for developers, protection for kids and workers | Future of Life Institute’s Pro-Human AI Declaration (March 2026); the Pro-Human AI Coalition (September 2026), co-founded with Project Liberty and Ergo Impact, with Omidyar and Mila on the steering committee | The phrase that won 2026. Safety science goals, Counterweights language, populist tactics, and new money behind all three. See below |
| People-centered AI | Put people’s agency and rights first; fund civil society, not labs | Humanity AI coalition (Ford, MacArthur, Omidyar, Doris Duke and others) | Legacy foundations, present-day harms, pooled money |
| People-first AI | Help local nonprofits adopt AI; “access is agency” | OpenAI Foundation’s $50M People-First AI Fund (June 2026) | Lab foundations. The same “people” word as Humanity AI, pointed at adoption rather than accountability |
| Public interest AI / Public AI | AI built, owned or governed outside the profit motive; public alternatives to the labs | Omidyar, Current AI, Ford, MacArthur; Mozilla and the Public AI Network for the “public AI” version | Public builders. “Public AI” means public infrastructure, not just public benefit |
| Human-centered AI | Design AI around human needs and oversight | Stanford HAI and the academic mainstream; Project Liberty’s “human-centric digital infrastructure” | The most neutral of the set. Industry can sign on, so it tells you least |
| Humane technology | Products should not exploit attention, kids or mental health | Center for Humane Technology (Tristan Harris, Aza Raskin) | The social media fight carried into AI; now a core member of the pro-human coalition |
| AI for good / public purpose | Use AI to deliver on health, climate, journalism and crisis response; “AI must function as civic infrastructure” | Patrick J. McGovern Foundation ($75.8M across 149 grants, December 2025); Google.org; the UN’s AI for Good | Adoption and nonprofit capacity. Closest to State capacity in spirit, rarely political |
| Human flourishing | Judge AI by its effect on well-being, relationships, meaning and faith | Gloo’s Flourishing AI benchmark; Templeton-funded flourishing research; the Pro-Human AI Coalition’s own tagline | The bridge word to faith and conservative audiences. Watch for it in anything trying to reach the right |
| Democratic AI | Two meanings. Labs: US-led AI as “a clear alternative to authoritarian versions of AI.” Civil society: AI that is governed democratically | OpenAI for Countries; the Collective Intelligence Project and deliberation researchers for the second meaning | The clearest case of one phrase, two camps. Ask which one before you nod; the second meaning is built by the participatory AI crowd below |
| Participatory AI | Let the public set the rules for AI through assemblies and deliberation, and use AI to run the deliberation itself | BLOOM’s Public Assembly on AI (up to 20 civic hosts, local assemblies February to July 2027, aiming at state and national assemblies in 2028); Unify Montrose in Colorado; Change.org’s $100M, three-year rebuild around AI for local democracy, paid for by its own contributors | The civic-tech wing. Small money except Change.org, and the place where the second meaning of “democratic AI” gets built |
“Good for people” by camp
- Counterweights: not controlled by companies
- Public builders: built outside the companies
- Safety science: humans stay in charge
- Lab foundations: access to the tools
- State capacity: services that work
- Discovery: faster cures and growth
Who uses “pro-social”
- Omidyar: “tilt the playing field toward pro-social outcomes”
- Social cohesion field: Schirch’s 2025 Blueprint on Prosocial Tech Design Governance; New_ Public, Build Up, Integrity Institute, CHT as founding members
- Berkeley CHAI: the Prosocial Ranking Challenge, testing feed algorithms for polarization
- Plurality circle: Weyl, Tang and Mchangama, “Building a Prosocial Media Ecosystem,” Noema, May 2025
Who uses “pro-human”
- FLI: wrote the declaration; ran the $8M “Protect What’s Human” campaign in Iowa, Kentucky, Maine, Michigan and North Carolina
- Project Liberty: Frank McCourt’s money and staff; the coalition’s CEO and executive director came from Project Liberty
- Omidyar: steering committee seat
- Mila, Center for AI Safety: steering committee and member
- Labor and faith: AFT, AFL-CIO Tech Institute, SAG-AFTRA, Institute for Family Studies among the signatories
- Humans First, Irreplaceable: the newer organizing groups; Amy Kremer and Phil Aroneanu spoke at the Assembly
The pro-human coalition is the first vehicle built across camps
FLI launched the Pro-Human AI Declaration in March 2026 with signatures from Yoshua Bengio, Susan Rice, Steve Bannon, Mustafa Suleyman, Ralph Nader and Daron Acemoglu. By September it had become a nonprofit coalition of 50+ organizations, co-founded by FLI, Project Liberty and Ergo Impact, with Omidyar and Mila on the steering committee and the Center for AI Safety, the Center for Humane Technology, Encode, Minderoo and Public AI as members. Its workstreams run from a “pro-human AI stack” to policy deliverables for the June 2027 Geneva AI Summit, and its members claim bipartisan AI laws in Utah, California, Ohio, Connecticut and South Dakota. That is Safety science money, Counterweights language and populist tactics in one organization. It also complicates the finding above that populists have no philanthropic base: through FLI and McCourt, they now have one.
Critics read the brand differently. Timnit Gebru’s response to the declaration: “Who is in this alliance you say? Well none other than Steve Bannon.” Reporting on the September assembly noted FLI’s own funding roots (Elon Musk’s founding $10M, Jaan Tallinn on the board, about $1M from Coefficient Giving) and that the only concrete harms discussed on stage were chatbot harms to children.
Deep dive: AI and democracy
This is where Counterweights and State capacity overlap most. The central document is Packard’s March 2026 report, AI and Democracy: Perspectives from an Emerging Field, co-authored by Kelly Born, who leads Packard’s Democracy, Rights, and Governance work.
AI is an accelerant, amplifying both strengths and weaknesses across existing systems.
Shevin, Born and Kirby, “AI and Democracy: Perspectives from an Emerging Field,” Packard, March 2026
What the report asks funders to do
- Build up the institutions that will absorb AI: agencies, regulators, unions, newsrooms.
- Pay for independent testing and measurement, instead of trusting industry’s numbers.
- Take on who owns AI through competition, labor and corporate governance work.
Where it sits
The report focuses on institutions rather than on the technology itself. The authors note their interviews lean toward near-term-risk funders, so this is the Counterweights and State capacity view rather than the whole field.
What the report leaves out
The report maps 130+ organizations, including safety groups like METR and Epoch. It lists no campaign or electoral infrastructure groups. Its elections category covers administration and integrity only.
What comes next
Still researching
- Omidyar’s AI Collaborative grants
- Who funds deliberative tech and antitrust work
- Hewlett’s $19M democracy grants
- Conservative foundation grants to AI think tanks, by name and amount
- Who funds the Pro-Human AI Coalition, and how much
Watching
- Who will run Anthropic’s fund
- Humanity AI open call awards
- How Ford frames AI in its new strategy
- The pro-human coalition’s Geneva AI Summit deliverables (June 2027)
- Cities Deliver first awards, April 2027 (Round 1 proposals due November 29)
- ImpactOpen’s credits column: how much of the AI wave arrives as compute rather than cash
To write
- A profile for each camp and bridge funder
- Migration stories, named and dated
- The people layer: Tisné, Bengio, Princeton CITP, the Packard report’s authors
- The philanthropy decoder, with the pro- words, as its own post
Key sources
- Packard, AI and Democracy: Perspectives from an Emerging Field
- Longview, Extreme Power Concentration RFP
- Survival and Flourishing Fund, 2026 recommendations
- Manifund, AI Safety Funder Bulletin
- Global Fund for a New Economy
- Current AI launch
- European AI & Society Fund
- Inside Philanthropy, the Navigation Fund
- Nan Ransohoff, The third wave of American philanthropy
- Nick Martin, The $100 Billion Argument: Part 1, Part 2, Part 3
- ImpactOpen, TechChange’s index of open AI-for-good funding
- Roy Bahat, How to Actually Give Away Your AI Money
- Future of Life Institute, The Pro-Human AI Declaration
- Pro-Human AI Coalition launch, September 2026
- Omidyar Network, Our Perspective on Responsible Tech
- Lisa Schirch, Blueprint on Prosocial Tech Design Governance
- Weyl, Tang and Mchangama, Building a Prosocial Media Ecosystem
- Foundation for American Innovation, supporters
- Horizon Institute, about and funders
- CSET, $100M+ grant agreement
- Institute for Progress, funders
- Encode, funders
- SF Standard, OpenAI’s subpoenas of its critics
- The Block, Buterin on FLI’s political pivot
- Politico Magazine, How the AI Doomers Won 2026
- SF Standard on Humans First and its CAIS loan
- Protos on Irreplaceable
- Pahlka, Announcing the Recoding America Fund
- Knight, $50M for research on technology and democracy
- New York, Empire AI expansion
- CalCompute, history and status
- BLOOM, Public Assembly on AI
- Change.org, $100M for local democracy
- Rockefeller, Cities Deliver