ChatGPT Captures 88% of Identifiable House AI Spending

OpenAI's ChatGPT accounted for roughly 88 cents of every dollar the U.S. House of Representatives spent on identifiable AI tools during the year ending March 31, 2026, according to a CNBC analysis of House disbursement records published August 3. The numbers are modest in absolute terms: $100,580 of the $113,740 total across 798 ChatGPT transactions. Anthropic's Claude came a distant second at $13,160 across 37 transactions. But they represent the first clear snapshot of which AI vendors are ga

2 min
ChatGPT Captures 88% of Identifiable House AI Spending

OpenAI's ChatGPT accounted for roughly 88 cents of every dollar the U.S. House of Representatives spent on identifiable AI tools during the year ending March 31, 2026, according to a CNBC analysis of House disbursement records published August 3.

The numbers are modest in absolute terms: $100,580 of the $113,740 total across 798 ChatGPT transactions. Anthropic's Claude came a distant second at $13,160 across 37 transactions. But they represent the first clear snapshot of which AI vendors are gaining traction inside the legislative branch as Congress weighs how to regulate the industry.

ChatGPT purchases appeared in at least 71 House member offices, roughly one in six. Democratic offices dominated the spending, accounting for $54,165 in AI purchases compared to $15,782 from Republican offices. ChatGPT showed up in 44 Democratic offices and 27 Republican ones, while Claude appeared in five Democratic offices and one Republican office.

The partisan gap creates an awkward dynamic for Democrats, many of whom have warned about AI risks to workers, privacy, civil rights, and elections while their offices lead visible adoption.

What Staffers Are Using It For

Congressional staff are deploying ChatGPT and other AI tools across a range of legislative work: summarizing bills, drafting memos, preparing hearing materials, responding to constituents, sorting through policy research, and writing social media posts.

Rep. Julie Fedorchak (R-N.D.) told CNBC her staff built a searchable database of legislation and constituent materials using ChatGPT, then used it to generate briefing memos. The setup is "saving dozens and dozens of hours every week of staff time," she said.

Marci Harris, CEO of the POPVOX Foundation, which trains Congressional staff on AI use, noted that AI is also exacerbating the workload problem. Lobbyists and constituents use the same tools to send longer, more frequent material to offices already operating with fewer staff than in previous years.

What the Numbers Miss

The figures almost certainly understate total Congressional AI adoption. The analysis is limited to House disbursement records that explicitly name an AI vendor. It excludes free versions of ChatGPT and Claude, AI features bundled into broader software contracts such as Microsoft Copilot, and most Senate spending. Some offices may also route AI purchases through shared institutional accounts that are harder to trace.

The data also predates the June 2026 flurry of activity around AI regulation. OpenAI CEO Sam Altman met with lawmakers on Capitol Hill on June 3, and the company has since moved to expand government access to its models.

Vendor Stakes

The spending race matters because regulatory decisions are being shaped by lawmakers whose staffs are already embedded in one vendor's ecosystem. OpenAI, Anthropic, Google, and Microsoft are competing for influence across Washington as federal agencies decide how AI should be purchased, used, and governed.

For now, the visible numbers show a lopsided market. ChatGPT's 88 percent share and 96 percent transaction dominance suggest a default choice, not a competitive evaluation.

Sources

CNBC: ChatGPT dominates early AI spending in Congress as lawmakers weigh regulation — https://www.cnbc.com/2026/08/03/openai-chatgpt-anthropic-congress-house-ai-spending.html

TechCrunch: Congress's favorite AI tool? ChatGPT — https://techcrunch.com/2026/08/03/congresss-favorite-ai-tool-chatgpt

Written by

More to read

  • Vector Quantization for LLM Weights in Production: Comparing QuIP#, AQLM, and VPTQ Architecture, Dequantization Kernels, and 2-Bit Serving Economics

    Vector Quantization for LLM Weights in Production: Comparing QuIP#, AQLM, and VPTQ Architecture, Dequantization Kernels, and 2-Bit Serving Economics Scalar post-training quantization methods such as GPTQ and AWQ have become the standard for compressing large language models to 4-bit integer formats (INT4). At 4 bits per parameter, scalar techniques preserve over 98% of baseline 16-bit floating-point (FP16/BF16) model accuracy across common benchmarks. However, pushing scalar quantization below

    1 min
  • OpenAI Integrates GPT-5.6 Family into AWS Kiro with Reported 82% Cost Drop

    OpenAI has made its flagship GPT-5.6 model family available within Kiro, the spec-driven software development environment developed by Amazon Web Services. The release brings OpenAI's frontier reasoning and coding tiers, including Sol, Terra, and Luna, directly into AWS's agentic engineering platform. According to joint evaluations conducted by AWS and OpenAI on Terminal-Bench 2.1, executing complex software engineering tasks with GPT-5.6 Terra inside Kiro reduced total token expenditures by ro

    1 min
  • NVIDIA Enters Full Production on Groq 3 LPX, Hitting 3,400 Tokens per Second in Benchmarks

    NVIDIA has moved its Groq 3 LPX dedicated inference accelerator into full commercial production. Announced at Hot Chips 2026, the rack-scale accelerator system is designed as a purpose-built extension for NVIDIA's Vera Rubin NVL72 data center platform, targeting the compounding decode latency bottlenecks created by multi-step autonomous AI agents. European neocloud provider Nebius Group N.V. has committed as the first cloud infrastructure customer to deploy the accelerators, integrating them in

    1 min