OpenAI has started offering a version of ChatGPT designed for US government agencies. ChatGPT Gov includes many of the same features as the enterprise version of the chatbot, including access to the company’s GPT-4o model. “By providing our products to the U.S. government, we strive to ensure that artificial intelligence serves the national interest and the public good in accordance with democratic values, while empowering policymakers to responsibly integrate these capabilities to provide better services to the American people,” OpenAI said in a blog post published on Tuesday.
To date, US government officials have already used ChatGPT in their daily work. According to the company, since 2024, federal, state, and local government employees in 3,500 agencies across the country have sent more than 18 million messages. With today’s announcement, these same agencies can now host ChatGPT themselves in their own commercial or public cloud environment on Microsoft Azure. In practice, this should make it easier for government IT managers to ensure the secure use of the tool.
OpenAI’s decision to offer ChatGPT Gov comes after the company announced a partnership with SoftBank to build a $500 billion artificial intelligence infrastructure in the United States over the next four years. Many people immediately questioned whether OpenAI had the money to fund Stargate to the level it had announced. After the announcement, The Information reported that SoftBank and OpenAI would allocate about $19 billion to launch the project, which is much less than the $100 billion they promised to allocate “immediately.”
More broadly, the emergence of ChatGPT Gov comes amid uncertainty about the strength of the US artificial intelligence industry. On Monday, shares of US tech companies plummeted after DeepSeek, an AI assistant from a Chinese startup, dethroned ChatGPT as the top free app in the App Store over the weekend. DeepSeek allegedly spent less than $6 million to develop its R1 model.
