Artificial intelligence companies OpenAI and Anthropic have invested heavily in the 2026 U.S. congressional primaries, aiming to influence which candidates advance to Congress amid growing public concerns about AI regulation and its societal impacts. According to a review of federal campaign finance data, political groups aligned with the two companies have spent approximately $55.7 million on campaign advertisements and voter outreach, with the vast majority of these expenditures concentrated in primaries for safe seats unlikely to change party control.
The spending has primarily supported Democratic candidates, with about $28 million backing 21 Democrats compared to nearly $19 million supporting 30 Republicans, and candidates favored by these groups have generally seen electoral success. Both OpenAI-aligned and Anthropic-aligned political action committees (PACs) have backed candidates who are perceived as amenable to their agendas, sought to defeat those considered hostile to AI interests, and cultivated ties with front-runners expected to shape future AI-related policymaking.
Despite the significant financial involvement, the ads purchased by these AI-linked groups rarely mention artificial intelligence or data centers explicitly. Out of 95 unique ads analyzed, only seven directly referenced AI or AI companies, and none mentioned data centers. The messaging focused instead on broader themes such as innovation, public safety, and community support, sometimes promoting contrasting positions on divisive issues like immigration across candidates funded by the same PAC.
The two companies have notably divergent policy preferences: Anthropic generally advocates for stricter AI regulation, emphasizing safeguards to protect children, workers, and families, while OpenAI prefers fewer restrictions to accelerate AI development and maintain U.S. technological leadership. However, both share the overarching goal of ensuring that AI remains a central focus in policymaking and shaping Congress with favorable lawmakers. Representatives of both PACs have denied having shared agendas, with Anthropic’s spokesman emphasizing their commitment to imposing meaningful rules on AI and countering Big Tech’s political influence, whereas OpenAI’s spokesman described their efforts as building a coalition of informed leaders to balance innovation with responsible oversight.
Examples of AI-aligned spending include a $1.6 million investment by Anthropic’s Jobs and Democracy PAC in support of Representative Valerie P. Foushee of North Carolina, who subsequently won a narrow primary victory. Foushee, who serves on the House Science, Space and Technology Committee and a bipartisan AI task force, has advocated for AI regulations prioritizing public interest and accountability. Conversely, in states such as Utah and New York, AI PACs have targeted candidates seen as opponents of AI or proponents of strict regulations, such as Assemblyman Alex Bores in New York, who narrowly lost a primary amid campaigns funded by competing AI interests.
This financial involvement extends across party lines and often into districts with minimal general election competition, focusing on shaping the primary outcomes where candidate selection can determine future regulatory directions. Some political observers and candidates have criticized this approach as a strategic yet opaque effort to exert influence through proxy political spending, with communications experts warning that the AI lobbying presence is becoming increasingly difficult to track, paralleling established advocacy groups known for complex funding channels.
Looking ahead, political operatives anticipate that AI industry interests may alter their strategies in response to heightened public skepticism, potentially channeling funds through major party PACs or smaller, less transparent groups. This would mirror tactics seen in other influential lobbying efforts, further complicating transparency around the AI sector’s role in U.S. elections and policymaking.