by Ben Baldi, SVP of Global Public Sector at Tricentis 
Tricentis was proud to sponsor the Potomac Officers Club’s 2026 Digital Transformation Summit, which brought together leaders from across government and industry for timely discussions about modernization, artificial intelligence adoption and mission delivery.
One theme came through clearly across keynotes and panel discussions alike: digital transformation is not just about adopting new tools. It is about creating the conditions for those tools to succeed. That means strong requirements. Trusted data. Rigorous testing. Clear accountability. And, above all, a firm understanding of the mission needs technology is meant to serve.
My colleague, Brent George, strategic SAP Solution Extensions sales executive for Tricentis, moderated an engaging afternoon panel, “From Policy to Practice: Driving Digital Transformation and Responsible AI Across the Federal Enterprise.” The conversation featured Andrea Brandon, deputy assistant secretary for budget, finance, grants and acquisition at the Department of the Interior; Anil Chaudhry, senior adviser for the Department of Transportation’s Highly Automated Systems Safety Center of Excellence; and David Hinchman, federal senior executive for cybersecurity and IT modernization at the Government Accountability Office.
What Is the Role of Cross-Functional Collaborations in AI Deployment?
During the session, George encouraged panelists to share an honest and practical look at what it really takes to move AI from pilot to impact. One major takeaway was that AI success starts with cross-functional collaboration.
At DOI, for example, Brandon shared that project teams bring together stakeholders from the budget, finance, and chief information officer offices and other relevant functions early in the process.
That kind of coordination helps agencies address security, compliance and operational requirements from the start rather than trying to bolt them on later. It also improves the likelihood that the use case will actually succeed once deployed.
How Critical Is Requirements Gathering in Technology Initiatives?
Another clear theme was the importance of requirements discipline. In government, that has always mattered. In AI, it matters even more.
Agencies must be able to separate “need to have” from “nice to have,” and they must be clear about the problem in front of them and the right solution to address it. Too often, enthusiasm for the technology can outpace clarity around the requirements. As panelists noted, many technology initiatives fail not because the tools themselves are ineffective, but because the underlying requirements were never fully defined.
Why is Rigorous, Automated Testing & Validation Essential for Building Trustworthy AI?
Testing was one of the most important themes of the day, and for good reason. As AI evolves at a rapid pace, traditional policy, validation and review cycles are struggling to keep up. That reality makes a stronger case for modern testing approaches that are continuous, measurable and built to scale. Panelists spoke to the importance of evaluating outputs through clear performance indicators, including false positives, false negatives and comparisons between machine-generated and human-generated work. For agencies looking to build trust in AI, that kind of rigor is not optional. It is foundational.
But just as importantly, validation cannot be treated as a one-time event. AI systems are dynamic, and models, data and use cases change often. If agencies want confidence in how these systems perform in production, testing has to evolve from a checkpoint into an ongoing practice. That is where automation becomes essential. Automated testing and continuous validation are what make it possible to identify issues earlier, reduce risk and move forward with greater confidence.
The discussion around data readiness reinforced another reality: quality outcomes depend on quality inputs. Agencies that want trustworthy AI results need to invest in data quality, standardization and ownership. A practical path is to begin with a clearly defined, high-value use case, identify the minimum data needed to support it and validate results continuously from there. That kind of disciplined approach helps organizations learn faster while maintaining control over risk. 
AI Only With Mission-Driven Purpose
More broadly, this all points to a principle that should guide every AI initiative: agencies should not pursue AI for AI’s sake. They should pursue it to solve a specific mission problem, improve a measurable outcome or strengthen a critical process. The panelists noted that in a market full of hype, the organizations making the most progress are the ones that stay grounded in mission need, operational reality and clear definitions of success.
Responsible transformation is not about moving slowly. It is about moving deliberately — with strong requirements, trusted data, continuous testing and the right balance of automation and oversight. For government agencies looking to scale AI with confidence, that is what it will take to modernize in a way that is both innovative and accountable.
