Advancing Trustworthy AI and Scaling Innovation in Government: Insights from Leading Experts
In recent discussions at the AI World Government event, key government leaders shared their strategies for harnessing artificial intelligence responsibly and effectively. With a focus on mitigating risks and promoting best practices, these insights highlight how federal agencies are leading the way in AI adoption and trustworthiness.
Driving Trustworthy AI at the Department of Energy
Pamela Isom, Director of the AI and Technology Office at the Department of Energy (DOE), emphasized the importance of developing AI systems that are reliable, ethical, and aligned with strategic goals. She has been instrumental in expanding AI applications across the DOE, focusing on areas like cybersecurity, fraud prevention, and life-saving initiatives.
“AI isn’t just about data—it’s about trustworthy outcomes,” Isom explains. Her team advocates for a holistic approach, integrating AI projects into broader strategic portfolios. This ensures that risk mitigation, data quality, and system reliability are prioritized from the outset.
A core part of her message is vigilance over data sources. “Having vast amounts of data doesn’t guarantee quality or representativeness,” she notes. The DOE actively consults international partners, academia, and industry to ensure AI systems produce trustworthy results. She highlights the transformative power of AI—its ability to do what humans do, but faster and often more accurately—and underscores the necessity of continuous monitoring post-deployment to maintain integrity.
Guided by Presidential Executive Orders
Isom’s efforts are aligned with recent federal directives, such as Executive Order 14028 on cybersecurity and Executive Order 13960 on trustworthy AI. To operationalize these policies, she developed the AI Risk Management Playbook—a practical guide for identifying system vulnerabilities, ethical considerations, and mitigation strategies throughout the AI lifecycle.
This playbook offers real-world examples, like addressing discrepancies between expected and actual accuracy, helping agencies proactively identify and resolve issues. While currently internal to DOE, plans are underway to share this resource with other federal agencies, fostering a broader culture of responsible AI.
Scaling AI in Government: Best Practices from the GSA
Anil Chaudhry, Director of Federal AI Implementations at the General Services Administration (GSA), shared insights on how federal agencies can effectively scale AI projects. With over two decades of experience in technology and security, Chaudhry emphasizes collaboration and industry partnership as cornerstones of successful AI deployment.
“The government landscape is vast, with every agency at different stages of AI maturity,” he explains. Common use cases include improving efficiency, reducing costs, enhancing response times, and ensuring compliance. One key recommendation is thorough vetting of commercial solutions against the enormous volumes of data agencies handle—petabytes and exabytes of structured and unstructured information.
Chaudhry stresses the importance of evaluating industry partners’ expertise, especially their experience with large datasets, trend analysis, and automation tools like robotic process automation (RPA). Equally critical is understanding the talent behind AI solutions. “You need to know if your partner has the right skills or access to the right talent,” he advises, because success hinges on the ability to evaluate and grow AI capabilities.
Funding and data access are also vital. AI projects often require unpredictable investments—flexibility in funding streams helps accommodate iterative testing and data refinement. Securing reliable access to authoritative data sources, including sensor data for IoT applications, ensures AI systems operate with accuracy and timeliness. Chaudhry recommends establishing data-sharing agreements early, with privacy considerations addressed upfront.
Finally, infrastructure planning is crucial. As pilots transition into full-scale deployments, agencies must reserve adequate data center capacity and plan for the management of increasing endpoints. Proper infrastructure ensures scalability without disruption.
Building a Future of Responsible AI
Both experts underscore that responsible AI development is a continuous journey. From risk mitigation and policy alignment to industry collaboration and infrastructure readiness, these best practices serve as a roadmap for government agencies aiming to innovate confidently.
To learn more about how federal agencies are shaping the future of AI, stay tuned to ongoing industry events and resources dedicated to trustworthy, scalable, and impactful AI solutions.
