Inside the Leaked US Government AI Platform: What ai.gov Aims to Do

The Trump administration is preparing a massive AI rollout across the federal government, centered on a new website called ai.gov . The project, managed by the General Services Administration (GSA) and its Technology Transformation Services (TTS) team, was uncovered in a GitHub repository that was later removed.

According to the leaked code, the site is planned to go live on July 4 and will serve as a hub for three main AI services:

A conversational chatbot capable of performing various tasks.

A "super API" that lets government systems connect directly to AI models from OpenAI, Google, Anthropic, and others.

A monitoring console that tracks AI usage across agencies, showing which tools are most popular.

The platform will primarily rely on Amazon Bedrock models that have received FedRAMP certification, ensuring they meet federal security standards. One exception is a model from Cohere, which has not yet been certified.

The initiative is driven by Thomas Shedd, the new head of TTS, who previously worked as a software integration manager at Tesla. Shedd’s goal is to automate many tasks currently performed by federal employees and to adopt an "AI‑first" strategy similar to a tech startup.

However, the leak has triggered widespread skepticism. Critics point out the security risks of handling massive amounts of confidential data, the potential for code vulnerabilities, and the challenge of integrating AI with legacy government systems. Some observers worry the project may become a costly showcase rather than a functional tool.

Public reaction on social media ranges from concerns about privacy and security to doubts about the government’s understanding of AI. The success of ai.gov remains uncertain until the site actually launches.

Reference links: https://github.com/gsa-tts-archived/ai.gov