Anthropic reported $1.97 million in federal lobbying expenses for the second quarter of 2026, up from $1.56 million in Q1, while OpenAI reported $1.2 million, up from $1.02 million. The filings show both AI labs expanding work on regulation, infrastructure, copyright, cybersecurity, export controls and other policy areas.
Anthropic and OpenAI both increased their reported federal lobbying expenses in the second quarter of 2026, according to disclosures posted by the US Senate's Lobbying Disclosure Act database on July 20. Anthropic reported $1.97 million, compared with $1.56 million in the first quarter. OpenAI reported $1.2 million, compared with $1.02 million in Q1.
What the filings show
Anthropic's disclosure covers a wide policy portfolio, including AI legislation and standards, cybersecurity, infrastructure, national security, export controls, model distillation, energy, health care, government procurement, workforce issues and oversight. It lists contacts with Congress and several executive-branch offices.
OpenAI's filing lists artificial intelligence, cloud infrastructure, cybersecurity, copyright and privacy among its lobbying issues. It also reports contacts with the House, Senate and White House.
The dollar figures are each company's reported quarterly lobbying expenses. They should not be read as spending allocated to any one bill or policy position: the forms group multiple issues under a single reported total and do not break the expense into issue-level amounts.
A larger Washington presence
Bloomberg Government reported that Anthropic, OpenAI and Waymo together spent a record $4.3 million on federal lobbying in Q2. It also reported that Anthropic spent more than $3.5 million in the first half of 2026, already exceeding the company's total for all of 2025.
The filings document a broader policy footprint for frontier AI companies as Congress and federal agencies consider rules for model deployment, infrastructure, intellectual property, national security and the workforce. They do not establish that either company secured a policy outcome.
Why the disclosures matter
For AI teams, the most useful signal is the range of issues now moving through the policy process. Model provenance, export controls, procurement requirements, energy capacity, copyright and cybersecurity are increasingly connected rather than separate compliance tracks. Organizations building on frontier models should preserve data lineage, vendor terms, evaluation records and deployment controls that can be mapped to changing requirements without reconstructing the evidence later.
Key Points #
- 1Anthropic reported $1.97 million in Q2 2026 federal lobbying expenses, up from $1.56 million in Q1.
- 2OpenAI reported $1.2 million for Q2, up from $1.02 million in the previous quarter.
- 3The filings cover broad portfolios and do not allocate the reported totals to individual bills, agencies or policy positions.
Scoring Rationale #
The disclosures show a measurable expansion of policy activity by two frontier AI companies, with direct implications for regulation, infrastructure, intellectual property, export controls and procurement.
Sources #
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.