Anthropic's Federal Lobbying Expenditure Reaches $3.53 Million in First Half, Nearly Tripling

Anthropic was founded by former members of OpenAI's safety team, initially positioning itself with a differentiated focus on "Constitutional AI" and safety alignment. Previously, its lobbying efforts were limited, with only $3.1 million spent in total for 2025. This recent surge directly responds to the Commerce Department's export control actions against the Claude series models.

The capital strategy aims to convert financing and valuation advantages into access rights in Washington, lobbying the White House, Commerce Department, and Treasury Department to secure more lenient conditions on model releases, data center operations, and defense procurement rules, while hedging against potential regulatory risks before an IPO.

This is comparable to the early lobbying expansions of Google and Meta during peak antitrust and privacy legislation periods, as well as OpenAI's own doubling of expenditures during the same timeframe. Anthropic is currently transitioning from a narrative focused on technical safety to one centered on policy defense and industry standard-setting.

Essentially, this reflects regulatory changes: when the government intervenes in the release of cutting-edge models in the name of national security, lobbying becomes a necessary capital investment to redefine the boundaries of rules and competitive thresholds.

ABAB News · Cognitive Laws

1. The tighter the regulations, the more expensive the lobbying

2. The safety narrative ultimately needs to buy policy access

3. Technical leadership is less important than the power to set rules