AI Industry PACs Clash Over Election Strategy

Silicon Valley’s attempt to shape federal AI policy has fractured into a civil war. One week before the midterm election, two distinct camps of industry-backed groups are spending heavily against one another. The first camp favors rapid development and uniform federal rules. The second camp argues for stricter safety measures and state-level authority.

Federal Election Commission filings through October 20 show the financial scale of this conflict. Groups aligned with faster development have spent roughly $8 million in general-election congressional races. Groups advocating for stronger safeguards have spent about $6.9 million in the same period. The numbers reflect a strategy that shifted from cooperation to competition.

The division is visible in specific districts. In Florida’s 25th Congressional District, the primary race highlighted the split. Think Big, an affiliate of the Leading the Future network, spent $310,000 to support Democrat Jared Moskowitz. Defending Our Values, a super PAC favoring stricter guardrails, spent nearly $857,000 to boost Republican Scott Singer. That group also spent $740,000 to eliminate one of Singer’s primary opponents. Moskowitz and Singer will now face each other in the general election.

The model for this spending was the cryptocurrency industry. In the 2024 cycle, Fairshake and its affiliated super PACs spent more than $130 million. CoinDesk reported that candidates supported by these groups won 53 of 58 races, a 91% success rate. AI investors replicated this structure in 2025. Leading the Future launched in August with over $100 million in commitments. Backers include Andreessen Horowitz, Greg Brockman, and Joe Lonsdale. Josh Vlasto, a former Fairshake staffer, co-leads the network, which operates both Democratic and Republican PACs.

Critics viewed the move as an attempt to control the legislative agenda. Sacha Haworth of the Tech Oversight Project described the industry’s plan to "bully lawmakers into submission." A rival network formed in response. Anthropic committed $20 million to Public First Action in February and added another $20 million in July. The company states its funds are limited to policy and education, though separate super PACs tied to the group have spent millions on candidates.

The most expensive clash occurred in New York’s 12th District Democratic primary. Think Big spent $8.3 million to defeat Assemblymember Alex Bores, who authored major state AI safety laws. The Anthropic-backed Jobs and Democracy PAC spent about $15 million to support him. Bores lost to Micah Lasher by a margin of 39% to 35%. The combined $23.3 million spent by these two groups represented about 40% of all primary spending by the two camps nationwide.

In the general election, the spending pace has slowed but remains focused. Safeguards-focused groups have spent $825,000 this fall opposing two Republicans who benefited from primary spending by the other camp. Jay Feely in Arizona and Aaron Flint in Montana each faced $412,500 in opposition from the Guardrails Alliance. That group also spent about $412,000 against Democratic Rep. Jimmy Gomez of California.

Most of the money has already been deployed. About 74% of the roughly $78.5 million spent on federal candidates this cycle went to primaries. Leading the Future’s network reported $26 million in primary spending versus $8 million in the general election. Safeguards-focused groups reported $32.1 million in primaries and $6.9 million in the general election.

Public opinion may influence the outcome. A CNBC All-America Economic Survey found that 52% of registered voters want companies to pause advanced AI development until Congress passes regulations. Only 32% favor moving ahead. Sixty-three percent oppose an AI data center within 10 miles of their home.

Some major donors are adjusting their strategies. The Brockmans initially committed $25 million to Leading the Future but dropped plans for another $25 million, according to The New York Times. Groups funded by Meta returned $25 million this summer, cutting its planned state-level push from $65 million to about $40 million. OpenAI CEO Sam Altman told Politico earlier this month, "Something's clearly not working."