Anthropic's Lobbying Spend Surges from $720,000 to $8.11 Million in Two Years: Inside the Washington Power Game of an AI Safety Standard-Bearer

Federal lobbying disclosures show that Anthropic's federal lobbying spending rose from $720,000 in 2024 to more than $3.1 million in 2025 and $8.11 million

In November 2026, a batch of federal lobbying disclosure filings showed that Anthropic had been paying Continental Strategy lobbying fees continuously since the second quarter of 2025, with cumulative payments reaching $740,000 by September 2026. Continental Strategy's main Washington operator, Alberto Martinez, is former Senate chief of staff to U.S. Secretary of State Marco Rubio.

That $740,000 is only part of the bill. According to a Crypto Briefing report based on federal disclosure filings, Anthropic separately hired Ballard Partners in April 2026, and Continental Strategy's partners happen to include the daughter of White House Chief of Staff Susie Wiles, a former employer of Ballard Partners. The two firms together collected $1.63 million in lobbying fees from Anthropic. If all lobbying spending on Congress and government departments this year is combined, Anthropic's total federal lobbying in 2026 has reached $8.11 million, compared with more than $3.1 million in 2025 and only $720,000 in 2024. A nearly 11-fold increase in two years makes Anthropic stand out for the pace of its lobbying spending growth in the tech industry.

Who Is Pushing the Door, and Which Door

To understand this spending, one first needs to see clearly who is being lobbied and what issues are involved. According to l0g.fr's analysis of the disclosure filings, Continental Strategy's contract took effect on April 24, 2025; in the first three quarters it charged Anthropic $90,000 to $110,000 per quarter, and from 2026 the fee rose to $150,000 per quarter. In addition to Alberto Martinez, the team also expanded with Carlos Trujillo (former U.S. ambassador to the Organization of American States) and Caroline Farr (former Treasury adviser), forming a channel covering the State Department, Treasury Department, Commerce Department, and National Security Council.

The institutional targets of lobbying are quite broad: the Senate and House of Representatives, the Executive Office of the President, the Treasury Department, the Commerce Department, the Office of Management and Budget, the National Security Council, the Department of Defense, and the State Department are all on the list. In terms of coverage, this is not point-to-point lobbying aimed at a specific piece of legislation, but rather the establishment of influence channels across the entire executive-legislative intersection.

At the issue level, according to Crypto Briefing, Anthropic's lobbying list covers federal government AI procurement, defense policy and national security AI applications, and restrictions on acceptable use of AI. Ballard Partners was brought in in April 2026, precisely after the boundary dispute between Anthropic and the Pentagon over AI use in military, intelligence, and surveillance scenarios surfaced.

Export Controls: Are Public Statements and Lobbying Direction Aligned?

Judging from disclosure filings and public records, Anthropic's public position on export control issues is quite clear. On March 6, 2025, Anthropic submitted policy recommendations to the White House Office of Science and Technology Policy, explicitly calling for stricter controls on Nvidia's H20 chips for the Chinese market, requesting additional resources for the Commerce Department's Bureau of Industry and Security (BIS), and promoting anti-smuggling agreements between governments. On July 23, 2025, when the Trump administration temporarily allowed H20 exports to China, Anthropic publicly criticized the move.

From this perspective, Anthropic's lobbying on export controls is aligned with its public statements. Lobbying itself is a legal and common way to participate in policy in the United States. But when a company's brand value depends heavily on a narrative different from that of AI companies that purely pursue commercial interests, large-scale political public relations actions themselves create friction at the image level.

The Pentagon Dispute: Another Side of Commercial Interests

Crypto Briefing's report specifically points out that the direct context for hiring Ballard Partners was the dispute that erupted between Anthropic and the Pentagon over AI supply chain risks and acceptable use issues.

This context reveals the dual-motive structure of Anthropic's lobbying behavior: on the one hand, on export controls, Anthropic supports imposing stricter technology blockades on competitors; on the other hand, on defense AI procurement, Anthropic wants to win more government contracts while seeking rules favorable to itself in drawing the boundaries of acceptable use. Both are direct expressions of commercial interests; the former merely happens to align with its safety position, while the latter appears more directly as a contest over market access.

From the perspective of government contracts, federal AI procurement is one of the largest certain growth markets in the coming years. Whoever has more say in the formulation of acceptable-use standards can influence the setting of procurement thresholds at the technical specification level.

Industry Structure: The Arms-Race Logic of Policy Competition

Anthropic's lobbying spending is not an isolated case, but its rapid growth and precise layout serve as an industry reference. The trend of tech companies increasing investment in Washington has a long history, but what is special about this round is that the lobbying targets have shifted from congressional legislation to policy-making by administrative agencies, and from general technology regulation to the technical details of specific AI chip export lists.

When major AI companies all begin to buy policy influence on a large scale, a structure similar to an arms race emerges: once one company's lobbying investment produces results, other companies face costs if they do not follow suit. This logic will drive the lobbying spending of the entire industry to continue expanding, proceeding in parallel with companies' technological competition in AI capabilities.

For downstream developers and enterprise users, the tightness of chip export controls directly affects which computing power developers in different regions can access, and at what cost major AI service providers can expand infrastructure. The drawing of defense AI procurement standards will send signals to the market about the boundaries of compliant AI capabilities, which in turn affects enterprise customers' compliance considerations when choosing AI suppliers.

For the competitive landscape, Anthropic's investment in lobbying networks reflects, to a certain extent, its intention to compete head-on with larger players such as OpenAI and Google on the track of policy influence. The latter two have earlier and larger lobbying layouts in Washington; Anthropic is using concentrated investment in precise relationship resources to make up for the gap accumulated over history. This is a reasonable catch-up strategy, but its cost is rising rapidly.

Forward-Looking Signals

The following judgments are analytical inferences based on existing disclosure materials and do not constitute statements of fact.

In the short term, the policy direction of H20 chip export controls is the most direct indicator to watch. If controls continue to tighten or even expand to more chip models, Anthropic's lobbying logic will be borne out at the market level; if policy loosens again, the effectiveness of its lobbying on export controls will be in doubt, and more spending may flow toward lobbying on defense procurement issues.

Second is the renewal of the Ballard Partners contract in the fourth quarter of 2026. The firm was brought in only after the Pentagon dispute became intense; if the dispute cools or negotiations enter a new stage, the contract structure may adjust accordingly, which would reveal which policy front Anthropic believes needs continued pressure.

The longer-term structural question is: as AI governance moves from principled discussion into the stage of specific technical standard-setting, lobbying between companies and regulators will move from behind the scenes toward more transparent institutional participation. For a company like Anthropic, whose core narrative is responsible AI, this is both an opportunity and a pressure—transparency exposes its lobbying behavior to more public scrutiny, while also providing a more credible policy endorsement channel for its public positions. How to find a balance between the two will become a core proposition for the next phase of governance strategy for leading AI companies.