The Silicon Valley AI political groups that spent the year lining up behind friendly candidates are now spending against each other. CNBC reported on October 27 that Leading the Future, Public First Action and Innovation Council Action backed rival candidates in at least six primary races where their preferred picks competed head to head. That is a sharp change from the early pitch, which cast the AI industry as a single bloc with a single legislative wish list.
The money is large enough to matter. Leading the Future received $125 million from donors that include OpenAI’s Greg Brockman and Marc Andreessen. Innovation Council Action, backed by former White House AI and crypto czar David Sacks, raised more than $100 million. Public First Action, which Anthropic funded with $20 million, raised somewhere between $75 million and $80 million depending on the report. Put the three headline figures side by side and they top $300 million in raised or received money, though that is not the same as money spent.
For anyone tracking the AI sector, this fight sits upstream of every model launch and chip deal. The rules Congress writes in 2027 will decide how labs train, ship and sell their systems. We covered the capital side of that race in our Anthropic IPO filing breakdown, and this piece picks up the political side.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What “Turning on Each Other” Actually Means
A word of caution on the framing first. The phrase “turned on each other” is an interpretive headline, not a quote from any of the groups. What the reporting does confirm is narrower and still striking: competing spending, and direct clashes in at least six primaries. CNBC’s account says the groups backed different candidates in the same contests, so ad dollars from one AI-linked group landed against candidates favored by another.
That distinction matters for readers who want to understand the story rather than repost it. The groups still share a general interest in how Washington treats artificial intelligence. They disagree about who should carry that interest into Congress, and about how much regulation the industry should accept. Calling it a civil war overstates the case. Calling it a family argument with eight-figure budgets is closer to what the filings show.
Beyond that, the sample is small. Six head-to-head primaries sit inside a much bigger field. POLITICO, citing reporting on AI- and crypto-linked super PACs, put total spending by those groups at $127 million across more than 100 primary contests in 2026. Six clashes out of 100-plus contests means the groups mostly avoided each other, and the collisions that did happen drew attention because the combatants were supposed to be allies.
The Three Groups at a Glance
Each group has a different donor base, a different political lineage and, by most accounts, a different theory of how to win a seat in Congress. The table below pulls together the figures confirmed in the reporting we reviewed. Gaps are marked as unconfirmed rather than filled in.
| Group | Named backers | Reported money | Leadership detail |
|---|---|---|---|
| Leading the Future | Greg Brockman, Marc Andreessen, Benjamin Horowitz | $125 million received | Co-led by Zac Moffatt and Josh Vlasto |
| Public First Action | Anthropic ($20 million), Dario Amodei ($1 million personally) | About $75 million per one report, $80 million per another | Not detailed in the reporting we verified |
| Innovation Council Action | Backed by David Sacks | More than $100 million raised | Organized by Taylor Budowich |
| Think Big (Leading the Future-funded) | Most funding from Leading the Future, per FEC records | More than $8 million spent against Alex Bores | Reviewed by Scripps News |
| All AI and crypto super PACs combined | Multiple | $127 million across 100+ primaries | Win rate near 90%, per POLITICO-cited reporting |
Two details deserve a flag. The Public First Action total is not consistent across outlets, so we show both figures instead of picking one. And Brockman’s role needs care: he is OpenAI’s president and a co-founder, but the reporting names him as a donor to Leading the Future. It does not say OpenAI the company wrote that check.
Leading the Future: $125 Million and a Long Reach
Leading the Future is the biggest single pot among the Silicon Valley AI political groups, at $125 million from donors including Brockman, Andreessen and Horowitz. Scripps News reported that spending tied to the group bought television ads, text messages and direct mail in 36 House and Senate races. That footprint is wide. It shows a group trying to shape many contests at once rather than concentrating on a few marquee fights.
The Think Big Channel
The sharpest example in the Scripps reporting involves Think Big, which spent more than $8 million opposing Alex Bores in a congressional race. Federal Election Commission records that Scripps reviewed showed Think Big drew most of its funding from Leading the Future. In plain terms, a group with its own name on the ballot-box ads is largely an arm of the larger operation. That structure lets the main committee keep some distance from individual attack campaigns.
The $8 million figure is worth sitting with. It is a single race, which shows how much one outside committee can put behind one target. When outside money reaches that scale, a candidate has a harder time controlling the message. That is the real lever these groups hold, and it is why lawmakers watch them closely.
Public First Action: Anthropic’s $20 Million Bet
Public First Action took a different route to the same arena. CNBC’s reporting says Anthropic gave the group $20 million, and CEO Dario Amodei added $1 million of his own money. The total raised is disputed. One report puts it near $75 million, while another says $80 million. When outlets disagree like that, the conservative reading is the lower number, and we treat anything above it as unconfirmed.
This donation lands in an odd spot for Anthropic. The company has been preparing a public listing, which we examined in our look at Anthropic’s giving pledges and the fine print on AI philanthropy. A company heading toward public markets usually avoids headline-grabbing political fights. Writing a $20 million check to a group that competes with other AI-industry money suggests Anthropic decided the policy stakes outweighed the optics.
The San Francisco Standard has also tracked how people who warn about AI risks are turning into an organized political force, a trend that helps explain why a safety-minded lab would fund a campaign vehicle at all. The reporting we verified does not detail Public First Action’s positions on specific bills, so we do not describe them here.
Innovation Council Action: The White House Alumni Network
Innovation Council Action brings a different pedigree. CNBC says Taylor Budowich, a former White House deputy chief of staff for Donald Trump, organized it, and David Sacks backs it. Sacks served as the Trump administration’s White House AI and Crypto Czar. The group raised more than $100 million, which puts it level with or ahead of Public First Action and not far behind Leading the Future.
The link to the Trump orbit gives this group something the others lack: proximity to the administration whose agencies will write much of the real AI policy. A donor who wants a friendly rulemaking process may see that access as worth more than a bench of House members. That is our analysis, and the reporting does not say it outright. It does help explain why three well-funded groups could pick different candidates and still all claim to represent AI interests.
Where the Money Collided: Six Primaries
The six head-to-head races are the core of the story. CNBC reports that in each, the groups’ preferred candidates competed directly, which put AI-industry dollars on both sides of the same ballot. The reporting we verified does not list all six contests, so we will not guess at names or winners.
Local coverage adds texture. Cronkite News and Arizona PBS examined how AI-linked money reached key Arizona races, a reminder that this spending shows up in specific districts with specific voters, not just in national filings. Local races are where outside money reaches voters most directly.
Why would allies collide? Three explanations fit the facts. First, donors disagree on policy, with some wanting a single national framework and others wanting stricter guardrails. Second, groups run on different timelines, and a committee focused on one candidate’s loyalty will not stand down because a cousin group likes someone else. Third, primary elections reward early money, so each group moves before it can coordinate. None of these requires a feud, and all of them produce the clashes CNBC describes.
The Scoreboard: $127 Million, 100+ Contests, About 90% Wins
POLITICO’s reporting gives the clearest scorecard. AI- and crypto-linked super PACs spent $127 million across more than 100 primary contests in 2026 and won about 90% of them. Divide the first number by the second and you get just under $1.3 million per contest on average. That is a rough average only, since the spending was uneven and the reporting covers both AI and crypto groups.
| Metric | Figure | Named source |
|---|---|---|
| AI and crypto super PAC spending, 2026 primaries | $127 million | Reporting cited by POLITICO |
| Primary contests covered | More than 100 | Reporting cited by POLITICO |
| Share of contests won | About 90% | Reporting cited by POLITICO |
| Primaries with direct AI-group clashes | At least 6 | CNBC |
| House and Senate races with Leading the Future-linked ads, texts, mail | 36 | Scripps News |
| Think Big spending against Alex Bores | More than $8 million | Scripps News |
| Anthropic gift to Public First Action | $20 million | CNBC |
| Dario Amodei personal contribution | $1 million | CNBC |
A 90% win rate sounds like proof of influence, but read it with care. Super PACs often back favorites, and a favorite wins most of the time with or without outside help. The figure shows the groups pick winners. It does not show they create them. Anyone who wants to credit the money for the results should ask what the same races would have looked like with no AI spending at all. No one has published that counterfactual.
What the Group Leaders Say
The on-record statements we can verify all come from Leading the Future. We found no confirmed quotes from Public First Action or Innovation Council Action leaders, so this section reflects one side of the argument and says so.
A group spokesperson described the mission to MarketWatch this way: “Leading the Future’s mission has always been to build a deep bench of pro-innovation champions who can work together in Congress to advance a strong, responsible national regulatory AI framework.” The statement appeared in a Morningstar-hosted MarketWatch report on the group’s midterm spending. Note the phrase “national regulatory AI framework.” It points toward a single federal standard, which is a specific policy position and not a generic call for innovation.
Co-head Zac Moffatt framed the group as an underdog. “I would argue we’re the David in this situation,” Moffatt said, according to a POLITICO Magazine profile, describing the group’s position against a wider network of nonprofits, PACs, lobbyists, media outlets and technology executives. A $125 million war chest makes the David label a stretch, but it tells you how the group wants donors and voters to see it.
Moffatt also treated November as a waypoint. “November is a stop on the journey. It’s part of a larger destination,” he said, per POLITICO’s Digital Future Daily. Co-leader Josh Vlasto echoed the urgency in a CNBC interview, saying, “It is so important that we do this now and urgently, because it is still the early innings of the technology, but it is being adopted quickly, at scale.” The interview is archived on a congressional press page.
Both men point to the same idea: the 2026 midterms are a first step. If that is right, the clashes in this year’s primaries are early skirmishes in a multi-cycle contest over who speaks for the AI industry in Washington.
Why Democrats and Tech Drifted Apart
Party alignment helps explain the donor map. Reid Hoffman said the Democratic Party “really did alienate” parts of the technology industry. That short remark carries a lot. It concedes that a sector once seen as a reliable source of Democratic money and talent has loosened its ties.
The groups reflect that drift. Innovation Council Action sits closest to the Trump orbit through Budowich and Sacks. Leading the Future and Public First Action present themselves in less partisan terms in the reporting we verified. The Yahoo Finance UK roundup of the biggest midterm donors shows how many corners of the tech and finance world are now writing checks, which makes a single “tech vote” harder to define than it was a decade ago.
Historical Context: Tech Learns the Super PAC Game
The technology industry used to lobby, not campaign. Trade groups and registered lobbyists handled Washington, and CEOs gave modest sums to candidates they liked. Super PAC money changed that math. Once outside groups could raise unlimited sums, any industry with a few dozen wealthy backers could field a political operation larger than most party committees in a given primary.
Crypto-linked groups showed AI donors what that looks like in practice, and POLITICO’s $127 million figure lumps the two industries together for that reason. The AI groups reportedly followed a similar playbook: pool money, fund committees with neutral names, buy ads in dozens of races. What the AI sector adds is internal division. Crypto’s political money mostly pointed one direction. AI money is pointing in at least three.
That division mirrors a split inside the industry itself. Labs disagree on open weights, on export controls and on how much safety testing should be mandatory. Our look at the Manhattan Institute’s safe harbor proposal for open weights shows how fast one policy idea can pull the industry in two directions, and the PAC map is the electoral version of the same argument.
Competitive Comparison: Three Strategies, One Industry
Comparing the groups on structure and posture shows why collisions were likely. The table sticks to what the reporting confirms and flags the rest.
| Dimension | Leading the Future | Public First Action | Innovation Council Action |
|---|---|---|---|
| Money | $125 million received | About $75 to $80 million | More than $100 million |
| Notable backers | Brockman, Andreessen, Horowitz | Anthropic, Amodei | Sacks (backer), Budowich (organizer) |
| Stated aim | Pro-innovation bench, national AI framework | Not confirmed in verified reporting | Not confirmed in verified reporting |
| Reach | 36 House and Senate races | Not confirmed | Not confirmed |
| Outside arm | Think Big, more than $8 million vs. Bores | None reported | None reported |
| Administration link | Not reported | Not reported | Budowich and Sacks, Trump alumni |
Leading the Future is the widest and the most structured of the three. Innovation Council Action has the closest ties to the executive branch. Public First Action carries the sharpest brand contrast, since an AI lab itself is funding it. Each group answers a different question: who wins seats, who shapes rules, and who speaks for the labs.
Market Impact: Labs, Listings and the Model Race
The political fight lands on a market that is moving very fast. One release tracker counted 150 model releases from 22 companies in October 2026, after 213 in September. Reports list OpenAI’s GPT-6.1 Terra on October 15, Google’s Gemini 4 Argon reaching general availability in the Gemini API on October 13, and Anthropic’s Claude Haiku 5.5 on October 7. For a head-to-head view of the current frontier, see our GPT-6.1 Sol vs Sonnet 5.5 vs Gemini 4 Argon comparison.
Policy risk now prices into that race. Labs that raise capital need to tell investors what rules might apply, and a Congress shaped by split AI money is harder to forecast than one shaped by a single lobby. Funding costs are already a pressure point, as our report on the AI cash crunch and Meta’s $95 billion ask showed. Add regulatory uncertainty and the cost of capital for smaller labs and neocloud operators goes up, not down.
Global readers should watch this too. The US sets the pace for export policy and procurement, and decisions made in a US committee room reach labs and customers abroad. Our analysis of the WAICO and Pax Silica blocs traces how those choices push other countries into camps. A more fragmented US industry position makes the bloc math harder for everyone else.
Five Predictions for the Rest of the Cycle
These are our expectations, based on the facts above. None comes from the groups themselves.
- Fewer direct clashes in November. General elections have two nominees, so the groups have less room to back rivals inside one party. Expect them to line up more often against the same candidates or on the same side of close seats.
- The 90% win rate becomes the main talking point. Every group will cite its record, and few will address the favorite-picking problem. Watch for reporters pressing the counterfactual question.
- Filings drive the next wave of stories. Scripps traced Think Big’s funding through FEC records. Expect more tracing of pass-through committees as post-election reports land.
- The fight moves to Congress in 2027. Moffatt called November a stop on a longer journey, and the real prize is a federal AI framework. Lobbying budgets will follow the campaign money.
- Lab-versus-donor distinctions get sharper. Anthropic gave corporate money while Brockman gave personally. Expect scrutiny of whether company policy and executive giving diverge, especially at firms nearing public markets.
What to Watch Before the General Election
Three things will tell you whether the split deepens. First, the final pre-election FEC reports will show whether donors keep funding all three groups or consolidate around one. Second, any new head-to-head races in the general election would suggest the groups cannot coordinate even when the stakes rise. Third, statements from Public First Action and Innovation Council Action leaders would fill the gap in this story, since we could verify on-record quotes only from Leading the Future.
Readers who follow the model market should track the policy calendar with the same care as release dates. For more on where the technology itself is heading, browse our AI and machine learning coverage. The political and technical stories now feed each other, and the next cycle of model launches will arrive in a regulatory climate these groups are helping to build.
FAQ: Silicon Valley AI Political Groups in the 2026 Election
Which Silicon Valley AI political groups are involved?
The reporting identifies Leading the Future, Public First Action and Innovation Council Action as the main AI-aligned groups spending in the 2026 election.
How much money has each group raised?
Leading the Future received $125 million. Innovation Council Action raised more than $100 million. Public First Action raised about $75 million in one report and $80 million in another, so the exact total is not settled.
Did the groups really fight each other?
CNBC reports they spent against one another in at least six primary races where their preferred candidates competed directly. The phrase “turned on each other” is a headline framing and not a quote from the groups.
What did Anthropic contribute?
Anthropic gave $20 million to Public First Action, and CEO Dario Amodei gave $1 million personally, according to CNBC.
Who is behind Innovation Council Action?
Taylor Budowich, a former White House deputy chief of staff for Donald Trump, organized it, and David Sacks, who served as the White House AI and Crypto Czar, backs it.
How successful have AI and crypto super PACs been?
Reporting cited by POLITICO says AI- and crypto-linked super PACs spent $127 million across more than 100 primary contests and won about 90% of them. That rate partly reflects backing favorites and does not prove the spending decided the outcomes.
What is Think Big?
Scripps News reports that Think Big spent more than $8 million opposing Alex Bores in a congressional race, and FEC records show it received most of its funding from Leading the Future.
