Political groups aligned with OpenAI and Anthropic have spent $55.7 million on the 2026 U.S. midterm elections, with the vast majority of the spending directed toward party primaries and congressional seats considered unlikely to change party control, according to a New York Times analysis of federal campaign-finance data.
Key Highlights
- Groups aligned with OpenAI and Anthropic have spent $55.7 million on the 2026 midterms, according to the New York Times analysis.
- About $52.6 million went toward safe congressional seats, with spending concentrated largely in primary contests.
- OpenAI-aligned and Anthropic-aligned groups have different approaches to AI regulation.
- The money is being spent through super PACs and affiliated political committees, rather than directly from the companies.
- An analysis of 95 political advertisements found that none mentioned data centers, while only some directly referred to artificial intelligence.
- The spending has been heavier on Democratic candidates than Republican candidates, according to the analysis.
- The campaign spending comes as Congress prepares to consider future rules governing artificial intelligence.
OpenAI and Anthropic Political Networks Spend $55.7 Million
The political networks aligned with OpenAI and Anthropic have emerged as significant players in the 2026 U.S. midterm elections.
According to the New York Times analysis, $52.6 million of the $55.7 million spent by the networks went toward so-called safe seats. These are congressional districts where the general election is generally less competitive, but where party primaries can determine which candidate ultimately holds the seat.
The spending has largely focused on primary elections rather than competitive general-election contests.
Leading the Future and Public First Back Different AI Policies
The spending is being carried out through political committees associated with two competing approaches to artificial intelligence regulation.
Leading the Future, which is aligned with OpenAI President Greg Brockman and backed by venture capital figures including Marc Andreessen and Ben Horowitz, has generally supported fewer restrictions on the AI industry.
On the other side, Public First Action, which has received major financial backing from Anthropic, has promoted stronger safeguards and greater transparency around artificial intelligence.
The two networks have established separate Democratic and Republican political vehicles through which their spending is conducted.
Most Campaign Ads Avoided Direct AI Messaging
Despite the scale of the spending, the campaigns have not generally made artificial intelligence the central issue in their advertisements.
The New York Times reviewed 95 advertisements purchased by four committees connected to the two networks. The analysis found that none of the advertisements mentioned data centers, while only a small number referred directly to artificial intelligence or AI companies.
The ads instead focused on broader campaign issues and candidate profiles, allowing the political groups to support or oppose candidates without necessarily making AI regulation the headline issue for voters.
Valerie Foushee Race Draws AI-Linked Spending
One race highlighted in the analysis was a Democratic primary involving Representative Valerie Foushee.
The Anthropic-aligned Jobs and Democracy PAC spent approximately $1.6 million supporting Foushee, who won the primary by less than one percentage point, according to the report.
The race illustrates how AI-linked political spending has extended into individual congressional primaries, where the outcome can determine which lawmakers eventually take part in debates over technology legislation.
Greg Brockman Withdraws Planned $25 Million Donation
OpenAI President Greg Brockman has also become a central figure in the political spending debate.
Brockman and his wife, Anna, previously donated $25 million to Leading the Future and had planned another $25 million contribution for 2026. The Wall Street Journal reported that Brockman later withdrew the planned second contribution following concerns about the political activity becoming a distraction.
OpenAI has maintained that such political contributions were made by individuals rather than directly by the company.
Brockman and other technology investors have also contributed to political committees supporting candidates and causes beyond AI regulation.
Anthropic Expands Political Spending
Anthropic has also increased its involvement in the 2026 political cycle.
The company has provided at least $20 million to Public First, while reporting has described a broader $40 million commitment to the network. Anthropic CEO Dario Amodei has also made personal political contributions connected to the effort.
Public First and its affiliated committees have supported candidates and policies associated with stronger AI safeguards and regulatory oversight.
AI Regulation at the Centre of the Political Spending
The political activity comes as Congress faces decisions about how the federal government should regulate increasingly capable AI systems.
The competing political networks have different positions on issues including federal and state regulation, AI safety standards, transparency and restrictions affecting the development and deployment of advanced AI systems.
The 2026 midterm elections therefore provide an important avenue for technology-industry donors and affiliated political committees to support candidates who may later participate in congressional debates over AI policy.
The New York Times analysis shows that much of that influence campaign has so far taken place in primary elections rather than in the most competitive general-election races. For more updates, follow us on X.
