OpenAI ends $1 federal deal, offers 50% discount to US agencies

September 10, 2026 at 06:15 PM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

OpenAI is ending its $1-per-year pilot program for US federal agencies and moving to a usage-based pricing system with a 50% discount from standard prices, the General Services Administration said. The new agreement takes effect October 1 and lasts more than two years. The discount will also extend to state, local, and tribal governments.

Key Facts

The General Services Administration said 3.5 million federal employees used OpenAI's ChatGPT during the $1-per-year pilot, generating $1.4 billion in cost savings.
The new 50% discount applies to OpenAI's models, including the GPT-6 Astra model, and takes effect on October 1.
OpenAI will announce Thursday that the discount expands beyond federal agencies to state, local, and tribal governments.
Rival Anthropic PBC negotiated the same $1-per-year price days after OpenAI's 2025 deal with the GSA.
Defense Secretary Pete Hegseth labeled Anthropic a supply-chain risk, and President Donald Trump ordered all agencies to stop using Anthropic.

Pricing Transition

OpenAI joined the GSA's OneGov marketplace in August 2025 to drive adoption of its models. The $1-per-year terms were the steepest tech discount the government's central purchasing arm had ever negotiated. The new usage-based pricing gives federal workers a 50% discount from standard prices. The agreement with the government takes effect on October 1 and lasts more than two years.

Anthropic Rivalry

Anthropic PBC negotiated a $1-per-year deal with the GSA days after OpenAI's 2025 agreement. Anthropic CEO Dario Amodei spoke at a Washington Union Station event to promote its AI tools to federal workers. A dispute with the Pentagon over surveillance and autonomous strike guardrails led Defense Secretary Pete Hegseth to call Anthropic a supply-chain risk. President Donald Trump ordered all agencies to stop using Anthropic, and the company sued the US government in response. The GSA initially directed removal of Anthropic from its central AI model platform but reversed course after a preliminary injunction from a California judge.