The Evolving Landscape of AI in Federal Agencies: Insights from Imagine Nation ELC 2023
HERSHEY, Pa. — The annual ACT-IAC’s Imagine Nation ELC conference has become a pivotal gathering for federal technology officials, and this year was no exception. However, unlike the previous year, which was dominated by discussions surrounding President Joe Biden’s groundbreaking artificial intelligence executive order, the 2023 conference reflected a more tempered approach. Attendees shared their experiences with AI, navigating the “hype cycle” that has emerged since the order was issued, and emphasizing the need for a balanced perspective on the technology’s capabilities and limitations.
Embracing the Fundamentals of AI
Zach Brown, the Chief Information Security Officer (CISO) at the Federal Deposit Insurance Corporation (FDIC), articulated a pragmatic approach to AI adoption within his agency. He emphasized a focus on “fundamentals” and “the basics,” which allows the FDIC to leverage AI for efficiency while ensuring that human oversight remains a critical component. Brown highlighted the use of an AI coding assistant, a generative AI tool that streamlines code creation and analysis. However, he cautioned that the effectiveness of such tools heavily relies on skilled developers who can interpret the outputs and communicate their implications to the workforce.
“It’s not magic,” Brown remarked, acknowledging the complexities of AI technology. He pointed out that while AI tools can enhance productivity, they must be approached with a clear understanding of their limitations. The hype surrounding AI, he noted, is likely to undergo a correction as the market adjusts to the realities of the technology.
The Ubiquity of AI Tools
The increasing availability of AI tools has transformed the landscape for federal agencies. Brown noted that “every single product” the FDIC acquires now incorporates some form of AI. Shelly Hartsook, acting associate director of the Cybersecurity and Infrastructure Security Agency’s (CISA) Cyber Capacity Building Subdivision, echoed this sentiment, observing a trend toward AI products that are more integrated and user-friendly. She likened the current environment to a well-equipped kitchen, where agencies can experiment with innovative solutions using the best tools available.
Hartsook emphasized the potential for “genuinely unique use cases” arising from this newfound accessibility to AI tools. However, she also cautioned against agencies developing independent solutions, advocating for a collaborative approach that leverages shared resources and expertise.
Experimentation and Automation
At the State Department’s Directorate of Defense Trade Controls, CIO Karen Wrege is keenly aware of the importance of experimenting with AI tools like Microsoft Copilot and GitHub Copilot. While her team is currently focused on addressing cybersecurity requirements, Wrege recognizes the promise of AI in automating manual processes. “A lot of these things we’re doing, we’re doing manually,” she stated, underscoring the potential for AI to enhance efficiency and improve code quality from the outset.
Similarly, Paul Blahusch, CISO at the Department of Labor, highlighted ongoing initiatives involving AI, including chatbots and coding assistants. He has encouraged his team to propose generative AI ideas that could streamline their work processes, particularly in updating internal policies in response to new regulations.
A Historical Perspective on AI
Greg Touhill, the first federal CISO and now director of the CERT Division at Carnegie Mellon’s Software Engineering Institute, provided a historical context for the current AI landscape. He noted that the advancements we see today are the culmination of decades of planning and development. “AI is an overnight sensation — over 65 years in the making,” he remarked, emphasizing the need for continued investment in the workforce to effectively apply AI technologies to various missions.
Touhill stressed the importance of aligning the right type of AI with the right mission to enhance national security and prosperity. As federal agencies continue to navigate the complexities of AI, his insights serve as a reminder of the long-term vision required to harness the technology’s full potential.
Conclusion: A Balanced Approach to AI
As federal agencies grapple with the implications of AI technology, the discussions at Imagine Nation ELC 2023 highlighted the importance of a balanced approach. While the excitement surrounding AI tools is palpable, officials are increasingly aware of the need for human oversight, collaboration, and a focus on fundamentals. The journey toward effective AI integration is just beginning, and as agencies continue to explore innovative solutions, the lessons learned from this year’s conference will undoubtedly shape the future of technology in the federal landscape.
About the Author
Matt Bracken is the managing editor of FedScoop and CyberScoop, overseeing coverage of federal government technology policy and cybersecurity. With a background in data-driven journalism, he has previously held editorial roles at Morning Consult, The Baltimore Sun, and the Arizona Daily Star. You can reach him at matt.bracken@scoopnewsgroup.com.