Published on: June 11, 2025 / update from: June 11, 2025 - Author: Konrad Wolfenstein
Leaked US ACI Initiative: Trump's comprehensive plans for AI.GOV from July 2025-Image: Xpert.digital
Leaved documents reveal Trump's ambitious AI-First strategy for the federal administration
Leaked github repositories show Trump's secret plans for government-wide AI integration
The Trump government is planning a far-reaching artificial intelligence initiative, which is due to start on July 4, 2025 and would establish a central platform called AI.GOV as the heart of the government-wide AI integration. These ambitious plans were publicly known by leaked Github repositories and staging websites before they were removed by the government. The initiative is under the direction of Thomas Shedd, a former Tesla engineer and ally of Elon Musk, and aims to implement a “AI-First strategy” in the entire federal administration.
Discovery and structure of the leaked plans
Github repository and staging website
The existence of the planned AI.GOV initiative was known for a briefly publicly accessible Github repository from the US General Services Administration (GSA). The repository disappeared from Github after media representatives asked questions, but archived versions remain available. The Technology Transformation Services (TTS) of the GSA work on this AI.Gov website, which is intended to serve as a central point of contact for government agencies to integrate AI into their processes.
An early version of the website hosted via Github revealed details about the project before it was also removed. The Staging website showed the planned start date on July 4 and the “Accelerate Government Innovation with Ai” (acceleration of government innovation with AI). This symbolic choice of the independence day as a start date underlines the national importance that the Trump administration of this initiative suggests.
Three-component system
The planned AI.Gov platform consists of three main components, which together are intended to form a comprehensive AI ecosystem for the federal administration. A chat bot should take on different functions, whereby the exact areas of application are not yet fully defined. This AI chat assistant is intended to help citizens and government employees with various inquiries and possibly reduce direct contact with authorities.
The second component is an “all-in-one-API” that would enable authorities to connect their systems with models from Openai, Google and Anthropic. However, code documentaries show that integrations with Amazon Web Services' Bedrock and Meta's Llama are also planned. This API strategy would give federal authorities access to the leading AI models in the private sector without negotiating each authority separate contracts.
The third element bears the name “Console” and is described as a groundbreaking analysis tool for authorities -wide implementations. This tool should enable authorities to monitor AI use in real time and to see which tools employees prefer. However, the surveillance function raises questions about data protection and employee monitoring.
Leadership and strategic vision
Thomas Shedd as a driving force
Thomas Shedd, who took over the management of the TTS at the end of January 2025, is the central figure behind this initiative. As a former Tesla manager, he brings experience as a software engineer and is described as an ally by Elon Musk. Shedd has communicated its vision significantly: he wants to lead the GSA like a software start-up and implement a government width, AI-First strategy for automation of many tasks that are currently being done by federal employees.
In joint audio recordings by meetings, Shedd explained its plans for “Ai Coding Agents”, which are to be made available for all authorities. His strategy includes the development of AI tools for fraud detection, the review of government contracts and to create a centralized platform for contract analysis. However, these automation plans are not without controversy, since cyber security experts warn that the government's automation is not comparable to the automation of a self -driving car.
Connection to Elon Musks Doge
The AI.Gov initiative is directly related to Elon Musk's Department of Government Efficiency (Doge), which has made the integration of AI into normal government functions. Shedd is described as a close ally of Musk, and his vision reflects DOGEs goals to rationalize and automate government processes. However, this connection also raises questions about conflicts of interest, especially in view of reports that Doges Team Musk's own Ki-Chatbot Grok tries to use in government agencies.
Doge employees supposedly urged the Department of Homeland Security to use GROK, although it was not approved for use in the authority. These practices could violate conflict of interests and endanger sensitive information about millions of Americans. Experts warn of a possible unfair competitive advantage for Musks XAI towards other AI service providers by using GROK in the federal government.
Technical implementation and security aspects
Fedramp certification and security standards
The Staging website indicates that the GSA works with Fedramp-certified providers. Fedramp (Federal Risk and Authorization Management Program) is a government program that offers standardized approaches for security assessment, authorization and continuous monitoring of cloud products and services. Most of the models listed in the API documentation are known to be fedramp-certified for government use. However, it was found that a model of the company Cohere appears in the API documentation without Cohere appearing to have Fedramp certification.
The GitHub documentary also shows that the Model Rankings website will publish, although the criteria for these reviews were not apparent. This ranking function could have a significant impact on the selection of AI providers by government agencies and raises questions about the transparency and fairness of the evaluation process.
Data protection and security concerns
Experts have expressed significant concerns about the widespread introduction of AI systems in the government, since these systems could process confidential data and personally identifiable information about citizens. The quick implementation of a AI-First strategy without adequate security precautions could create considerable security risks. These worries are increased by reports on Doges Access to strongly protected federal banks that contain personal information about millions of Americans.
Reports are particularly worrying that Doge employees have tried to receive access to emails from DHS employees and instruct them to train AI to identify communication that indicates that an employee is not “loyal” to Trump's political agenda. A group of about a dozen workers in a Department of Defense Authority has been informed by a manager in the past few weeks that an algorithmic tool monitors some of its computer activities.
Political and regulatory context
Lift of bidges regulations
The AI.Gov initiative is in the context of a broader deregulation strategy of the Trump administration in the field of artificial intelligence. Trump already raised a decree of his predecessor Joe Biden to regulate artificial intelligence. Bidens arrangement from 2023 aimed to reduce AI risks by obliging developers to disclose data. Instead, the Trump administration pursues an “innovation-friendly approach” that uses this technology to influence the future of government operations.
The Office of Management and Budget instructed government agencies to take on basic risk management protocols for important AI applications and to create a guideline for generative AI in the coming months. This instruction reveals two directives of the bid administration: one that obliged the authorities to implement protective measures for individual rights and ensure transparency, and another that aimed to impose restrictions on AI acquisitions.
Resistance at the state level
Despite the nationwide deregulative efforts, resistance is formed at the federal level. Prosecutors from 40 US states have called for congress to prevent a change in the law that would prohibit the states of all regulation of artificial intelligence for ten years. In a joint letter, they warned that "the consequences of a far -reaching moratorium would be considerable and that meaningful attempts to prevent known damage in connection with AI would completely destroy".
The prosecutors argue that this measure would affect “hundreds of existing and planned laws”, which were developed to protect against AI generated pornography, deeppake for disinformation and spam calls. These regulations have been developed over the years through careful considerations and, with the involvement of numerous actors such as consumers, industry and lawyers.
Security breakdowns and communication problems
The signal chat incident
In parallel to the AI initiatives, a suspected communication span of the Trump government caused turmoil. Jeffrey Goldberg, editor-in-chief of “The Atlantic”, reported that he was accidentally invited to a group entertainment of leading government officials via the Messenger app signal. The secret chat is said to have an impending attack on the Huthi militia in Yemen.
After the publication, a spokesman for the National Security Council confirmed that the chat course was most likely authentic and announced an internal exam. Defense Minister Pete Hegseth, who was mentioned as a participant in the entertainment, later denied the report, whereupon Goldberg accused him of the lie at CNN.
This breakdown illustrates the challenges of dealing with sensitive information in the new government and raises questions about information security that are particularly relevant because administration is also planning a comprehensive AI integration. Democratic politicians criticized behavior as “amateur -like” and one of the “most incredible injuries” of military secrets.
Leacted AI.Gov plans reveal Trump's ambitious digitization strategy for government
The leaked plans for the AI.Gov initiative reveal the extent of the Trump administration ambitions for AI integration in the federal government. While the initiative has the potential to modernize government services and increase efficiency, the hasty implementation, connections to private interests and the lack of transparency raise considerable concerns. The introduction planned for the 4th July will be an important test for whether the government can realize its ambitious AI goals without endangering the security and data protection of citizens.
The initiative is an example of the wider approach of the Trump administration to technology policy: Fast innovation and deregulation are in the foreground, while traditional security measures and supervisory mechanisms are reduced. It remains to be seen whether this approach will ultimately be successful, but the early signs indicate considerable challenges in implementation.
Suitable for:
Your AI transformation, AI integration and AI platform industry expert
☑️ Our business language is English or German
☑️ NEW: Correspondence in your national language!
Konrad Wolfenstein
I would be happy to serve you and my team as a personal advisor.
You can contact me by filling out the contact form or simply call me on +49 89 89 674 804 (Munich) . My email address is: wolfenstein ∂ xpert.digital
I'm looking forward to our joint project.