How scalable solutions are helping federal agencies unlock AI’s potential

The U.S. government has entered a new phase of data and innovation, aiming to maximize the potential of analytics and artificial intelligence. While progress is evident, three experts emphasize the need for improved infrastructure, streamlined data flows and robust governance to overcome the challenges of scaling AI from pilot programs to full production in a new video panel.

Bridget Carper, deputy chief information officer for architecture, engineering, technology and innovation at the Department of Energy (DOE), noted the enthusiasm for AI across agencies but highlighted hurdles such as budgeting, equitable access to tools and cybersecurity. “We didn’t budget two years ago for some of these tools,” says Carper. “Scaling access and ensuring compliance with data security standards remain significant challenges.”

Joel Meyer, president of public sector at Domino Data Lab and a former DHS official, echoed these sentiments, citing the urgency to deliver early successes amidst tight budgets. He emphasized the importance of systematic auditability in AI platforms. “Auditability at scale must be built into the platform,” says Meyer. “It’s not something you can do manually. It ensures consistency and effectiveness across complex processes.”

Olivia Zhu, assistant director for AI policy at the White House, underscored the critical role of governance and user-centric design in AI implementation. “Building guardrails into AI systems actually earns user trust and ensures effectiveness,” says Zhu. She also highlighted initiatives like the Chief AI Officers Council, which fosters knowledge sharing and governance across agencies.

The good news is that success stories are already emerging. DOE’s “Energy GPT” leverages internal data for productivity and decision-making, while the Navy’s Project AMMO has dramatically reduced AI model retraining times from six months to one week. “These practical applications show the transformative potential of scalable AI,” says Meyer.

Looking ahead, Zhu encouraged agencies to think big in their AI strategies. “What problems do you want to solve with AI? Work backward from that vision to build the necessary infrastructure and capabilities,” she says.

Carper added that listening to agency needs and aligning AI with business objectives is essential. “AI should solve the right problems, not just add complexity,” she says. Ultimately, with strategic planning, robust governance and early wins, federal agencies can unlock the transformative power of AI to serve the public better.

Learn how Domino Data Lab’s platform helps government agencies integrate AI into their missions rapidly, safely and cost-effectively.

This video panel discussion was produced by Scoop News Group, for FedScoop and underwritten by Domino Data Lab and AWS.