AI, Government, and the Future

March 2025

Author

Corner Alliance
March 31, 2025

AI, Government and the Future

Welcome to our weekly dive into the exciting world of Artificial Intelligence (AI) and its impact on the U.S. Government!

AI is progressing at an incredible pace, and we're just scratching the surface. With so much information out there, it can be overwhelming to keep up.

We're here to provide you with insightful analysis and a concise summary, delivered to you on a regular basis. Stay informed, stay up-to-date, and join us on this thrilling journey into the future of AI.

Spotlight

GSA Launches AI Chatbot to Boost Efficiency Amid Workforce Concerns - FedScoop

The General Services Administration (GSA) has introduced a new generative AI tool designed to enhance efficiency and automate repetitive tasks for their staff on Thursday, March 20th. This AI platform can answer simple user prompts and can assist with basic tasks including writing. GSA officials clarified that the tool is intended to foster a cultural shift towards AI utilization rather than to make official agency decisions.  The AI chatbot is isolated from the GSA's internal knowledge bases, and there are safety measures in place to prevent personal and sensitive information from being shared. While employee interactions with the chatbot are being logged, they will not be classified as federal records. Looking ahead, GSA plans to expand the tool's methodology for other agencies to utilize, and aims to see how the tool fits in the day-to-day workflows and make any adjustments needed.

New Guidelines Address AI-Powered Cyberattacks and Mitigation Strategies

The National Institute of Standards and Technology (NIST) has recently finalized its guide on combatting AI-powered cyberattacks and threats. These updates include an overview of the learning stages of generative AI models, a discussion on ongoing open problems in the field, and an index of various classes of attacks on different AI systems. As Artificial Intelligence and Large Language Models (LLMS) continue growing in popularity, so do related security threats to Federal agencies. The new NIST guidelines may serve as a starting point for other agencies’ own strategies and policies, though perhaps not a one-size-fits-all solution. While NIST’s finalized guidelines do mark a significant progress in AI security, there is much more discovery and work still to be done. The true test of cyber defense policies lies in the practical implementation and the ability to stay ahead of increasingly complex cyber threats.

DISA to Launch a GenAI-Based Chatbot Tool

The Defense Information Systems Agency (DISA) will soon be extending its classified network artificial intelligence assistant, SIPRgpt, to military personnel stationed in the Indo-Pacific Command region. This expansion represents a significant advancement in providing secure generative AI capabilities within classified environments. Operating on the Secure Internet Protocol Router Network (SIPRnet), SIPRgpt will enable military professionals to utilize advanced language model technology without requiring access to the commercial internet or to commercially-available tools. Once the tool is accredited, it processes sensitive data entirely within the classified network infrastructure, eliminating potential exposure to external commercial providers. Crucially, this tool is a part of a larger research-and-development effort by the Defense Department to determine the best path forward to utilize and deploy AI tools throughout its operations.