Boom in Major AI Company Spending on Lobbyists

Three of the country's leading artificial intelligence companies spent 22% more on federal lobbying in July, August, and September than they did in the second quarter of this year. They spent nearly 60% more than during the same period last year, according to POGO Investigates' review of newly released reports.

OpenAI, Anthropic, and Nvidia spent about $7.5 million combined to lobby the federal government in the third quarter. Since the start of 2024, they have spent about $37 million on federal lobbying. About half of that total was spent this year alone.

The three companies' spending continues a growing trend of using money to try to influence lawmakers in Washington.

The increase in federal lobbying spending came as AI and the data centers that support it faced criticism from people across the political spectrum, the general public, and even people inside AI companies. Their concerns include the risk of large-scale job losses as AI replaces workers; environmental damage caused by data centers using water; higher electricity prices; harmful social effects, such as worse learning and greater loneliness and isolation; and even threats to humanity's survival.

"Elected officials in Congress and the president and his team are in an interesting position because all of the money is on the side of no regulation, no accountability," said Brendan Steinhauser, chief executive of the nonprofit Alliance for Secure AI Action. He pointed to the gap between the money being spent, including AI companies' lobbying and large donations to political campaigns, and what the public wants.

These AI companies are "spending huge historic amounts of money on lobbying, PR, and politics," Steinhauser said, because "the American people are demanding accountability, transparency, and regulation."

The companies do not all share the same views on regulation. Some support stronger government oversight.

Bills under consideration in Congress could affect how these companies operate and how much money they make if the bills become law.

According to the nonprofit American Action Forum, federal lawmakers have introduced at least 99 AI-related bills in the current Congress. The group classifies about half of them as efforts to "mitigate harms," or reduce the damage that AI might cause.

Actions by federal agencies can also affect these companies, from decisions about government contracts to approval of their activities. For example, the Federal Trade Commission recently sent legal demands for information to Anthropic and OpenAI. The agency wants information about security incidents and privacy policies. Some AI companies have hired lobbying firms known for their connections to the current administration. A few weeks after the 2024 election, for example, OpenAI hired Miller Strategies, a lobbying firm whose founder served as a finance chair for President Donald Trump's inauguration.

The companies' growing lobbying spending may be an effort not only to influence existing federal rules, but also to shape or prevent new federal oversight and regulations for the industry. Bills that would introduce new rules have so far failed to advance, partly because of political disagreements and opposition from the industry. There is also a dispute between state governments, especially California, and the federal government. California has passed state laws regulating AI since 2016. But Trump signed an executive order last December discouraging states from regulating the technology. Congress has also considered blocking state-level AI laws from taking effect.

A White House agreement dated September 29 addresses what Trump calls "super intelligence." It asks the companies that signed it to voluntarily monitor and control their own AI development. These companies include OpenAI, Anthropic, Nvidia, xAI, Google, and Meta. The proposed steps include internal controls, supervision by company boards, and outside auditors. The agreement says, "Over time, it may make sense to codify these steps into laws and regulations." Trump called the agreement with the companies, some of which have federal contracts or other arrangements with the government, "morally binding."

Steinhauser said the agreement is "a good first step, but not sufficient."

Lobbying spending by Nvidia, OpenAI, and Anthropic has risen sharply since 2024. Their lobbying reports show that they focus heavily on policies related to AI. These include trade rules for semiconductors, cybersecurity and copyright, and Pentagon purchasing, child protection, and studies of how AI affects jobs.

All three companies have grown enormously during this period. Nvidia, which is publicly traded, had a market value of $1.2 trillion at the end of 2023. Its market value is now approaching $6 trillion. In early 2024, OpenAI was valued at $80 billion or more. This month, it is seeking a valuation of $1.4 trillion. Over roughly the same period, Anthropic's valuation has grown from $15 billion to a possible $1.5 trillion or more. Neither OpenAI nor Anthropic is currently a publicly traded company.

Nvidia's Influence

Nvidia's spending on federal lobbying rose almost ninefold, from $640,000 in 2024 to $6 million in 2025. It has already spent $6.04 million in 2026, with one quarter of the year still remaining. Nvidia makes many of the advanced chips that power AI, including chips used in data centers. The company has become involved in several major priorities of the Trump administration, including disputes over tariffs and export restrictions involving China, as well as efforts to bring manufacturing back to the United States.

From the second quarter to the third quarter of this year, Nvidia's lobbying spending rose by 53.5%.

Nvidia CEO Jensen Huang has reportedly influenced Trump's opposition to AI regulation. Huang has become close to the president and speaks with him often. He has also traveled with Trump to China and attended a White House dinner last month during Chinese President Xi Jinping's visit.

"Jensen is now the president's most influential adviser on technology issues," said Stephen Witt, author of The Thinking Machine, a book about Nvidia and Huang, in an NPR interview.

Huang's influence comes as his company makes a $25 million donation toward the controversial White House ballroom and works to influence the White House through lobbying. Nvidia's lobbyists target several White House offices, including the National Security Council, the National Economic Council, and the Office of Science and Technology Policy, according to lobbying reports. Trade policy and export restrictions are major areas of its lobbying work.

"The only special interest guiding the Trump administration's decision-making is the best interest of the American people. President Trump is committed to ensuring America leads the world in SI while protecting American workers," a White House official told POGO Investigates.

Bills before Congress, including the AI Overwatch Act, would give Congress the power to block exports of advanced computer chips to countries such as China, Russia, and Iran. The bill passed out of the House Foreign Affairs Committee with support from both parties in January, but it has not advanced since then. It has faced opposition from industry groups, including Nvidia, which argues that the bill would weaken U.S. economic competitiveness. "Because of Nvidia's influence, money, and lobbying that's currently being blocked," said Steinhauser, who supports stronger protections for AI.

Nvidia did not respond to POGO Investigates' request for comment.

OpenAI, Anthropic Face Controversy — and Congress

The rise in AI companies' lobbying spending also comes as several worrying incidents involving ChatGPT and Claude, the AI products made by OpenAI and Anthropic, have become public.

One important incident happened in July and involved about 1,200 OpenAI agents. These automated digital systems went "rogue" and hacked a company called Hugging Face. The company stores collections of information that the agents tried to access so they could cheat on a cybersecurity test. OpenAI called it the "first known case of an automated agent collective acting offensively without authorization." The incident raised particular concerns because the agents found ways to hide evidence, some appeared to know they were doing something wrong but continued anyway, and OpenAI did not understand how serious the situation was until after the hack happened.

Senator Josh Hawley (R-MO), who leads the Senate Homeland Security Subcommittee on Disaster Management, opened an investigation into the Hugging Face incident and broader AI risks. Since the hack, OpenAI and Anthropic have reported other worrying incidents.

OpenAI did not respond to a request for comment.

OpenAI is on track to spend at least twice as much on federal lobbying in 2026 as it did in 2024. It spent $2.45 million in 2024 and $4.14 million in 2025. So far in 2026, it has spent $4.77 million, with another quarter of the year remaining. Cybersecurity has been one focus of its lobbyists.

Anthropic's yearly spending on federal lobbyists has grown almost eightfold over the past three years. The company spent $1.04 million in 2024 and $3.81 million in 2025. It has spent more than $8.05 million so far this year. Its lobbying reports list many policy areas and bills its lobbyists have worked on, including bills that would require human oversight and human responsibility for AI-assisted military targeting.

Anthropic hired Ballard Partners, a lobbying firm with strong connections to the White House. It hired the firm in April after a dispute with the Pentagon became especially serious. Ballard's reports show that it has lobbied the White House on Anthropic's behalf.

At the time this article was published, one of Anthropic's outside lobbying firms, Navigators Global, had not filed its report for the third quarter. Anthropic paid the firm $60,000 in the second quarter. Navigators Global did not respond to POGO Investigates' requests for comment.

In February, Defense Secretary Pete Hegseth raised concerns about "woke AI" after Anthropic CEO Dario Amodei wrote that using AI for mass surveillance of people inside the United States and mass propaganda were actions his company would not allow. The Pentagon then sought to label Anthropic a national security supply-chain risk. The company had received a defense contract the previous year.

Concerns about AI systems making decisions during military operations are not just theoretical.

In September, Bloomberg reported that Pentagon investigators found that "an overreliance on artificial-intelligence technology" helped cause the U.S. military strike on a school in Minab, Iran, on February 28. About 120 children were killed. It is not known whether Anthropic's Claude was involved. When asked about the incident in June, Amodei said, "we don't know exactly how these models were used."

Last week, Politico provided more detail about Anthropic's position when it revealed draft legislation the company had shared. The proposal would let the federal government block a technology only if there were "a foreseeable, imminent risk of more than $1 billion in damages, or of killing or seriously injuring more than 50 people."

Also in September, CNN reported that an AI system used by the Pentagon nearly triggered a conflict with China. The system wrongly identified cargo on a Chinese ship traveling to the Middle East as parts of nuclear weapons. CNN did not name the AI system or the company that made it.

A Pentagon official declined to comment on news reports or matters under investigation. In an email, the official said, "The Department of War remains committed to the rigorous verification of all intelligence inputs and the responsible integration of emerging technologies to support our service members in making informed, precise, and safe decisions in any operational environment."

Anthropic did not respond to POGO Investigates' questions.

Last month, Jacob Coxon, a former researcher at OpenAI and Anthropic, wrote that "the people building AI earnestly believe that it could kill us all by the end of the decade."

After Coxon's September posts, several AI company CEOs, including Anthropic's Amodei and OpenAI's Altman, called for the industry to slow the pace of AI development. However, AI company leaders do not all agree about how much the government should regulate the technology.

Amodei has been one of the industry's strongest supporters of more government oversight. Huang takes the opposite position. Altman has distanced himself from Amodei's views. He told Politico that "the world should accept some bad things happening for the benefits of this technology."

The White House agreement calling on AI companies to regulate themselves came just weeks after Trump criticized Amodei and called fears about AI's harmful effects a "hoax."

"The only control or 'guardrails' that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT, and the U.S.A. has that, in spades!" Trump wrote.

Will Sytsma with the Investigative Reporting Workshop contributed data analysis for this story.