Governance and Context Determine AI Success
Discover why AI projects fail without governance, workflow integration, and strong data foundations, and how public sector leaders are building trusted AI systems.
Jonathan Alboum
Federal CTO of ServiceNow
Toolkit Shift: AI enhances technology leadership by delivering better services and uncovering new opportunities for efficiency.
Governance Necessity: Successful AI deployment requires early governance and integration into workflows to avoid silos and ensure visibility.
Human Oversight: AI aids operational decisions, but critical governance and strategic choices rely on human judgment.
Impact Strategy: An integrated AI strategy is essential for real impact, avoiding fragmented efforts that may not scale.
Data Quality: The quality of data underpins AI effectiveness; fragmented or outdated knowledge leads to unreliable outcomes.
Jonathan Alboum is the former CIO of the USDA and current Federal CTO of ServiceNow, which acts as a control tower for AI-driven transformation in government.
We sat down with Jonathan to learn how AI is changing technology leadership in the public sector. Here's what he had to say.
New Tools For An Old Challenge
I serve as the Federal CTO at ServiceNow, collaborating with federal agencies to deliver digital workflows that enhance experiences and unlock productivity. Throughout my career, I've consistently focused on leveraging technology to empower employees and customers, helping them become more effective in their jobs and deliver essential support services.
Want more from The CTO Club?
Create a free account to finish this piece and join a community of CTOs and engineering leaders sharing real-world frameworks, tools, and insights for designing, deploying, and scaling AI-driven technology.
Have an account? Log In
I began that work in government, serving in senior leadership roles, including the Deputy CIO and CIO for the Food and Nutrition Service in the Department of Agriculture, and later as the CIO for the entire USDA. I was also an executive at GSA, where I led the creation of its consolidated IT organization.
What’s changed over time is the toolkit. Today, AI builds on IT’s core expectations — delivering better services, enhanced experiences, and greater mission impact — while opening new opportunities for efficiency and transformation.
I have focused on applying these new tools to a familiar challenge: aligning technology, people, and processes to help agencies achieve their missions more effectively.
A Different Sort of Technology Organization
Our federal team at ServiceNow takes a different approach than a traditional technology organization. Rather than a centralized product development group, we have a team of solution consultants who work closely with our customers and account teams to deliver tailored outcomes.
They ensure our technologies and capabilities are well understood, architect solutions that work effectively within a specific agency or organization, and connect systems, people, and processes so the platform delivers real outcomes.
ServiceNow serves as the control tower for AI-driven transformation in government. We connect disparate systems and organizations, and use our platform capabilities, including AI, to drive outcomes. But that only works if you deploy with a clear architecture, strong governance, and alignment with mission goals.
Why AI Informs Operational Decisions, But Humans Lead
AI accelerates understanding of what’s happening and what we can do, but humans maintain ownership of what we should do — and under what guardrails.
We use AI for day-to-day operational decisions: intelligent incident routing, pattern surfacing in tickets, auto-generating knowledge articles, and suggesting backlog automation.
However, humans lead critical decisions like architecture, security reviews, and overall governance. These require judgment on risk, mission impact, compliance, and long-term strategy that models cannot fully capture.
So, AI accelerates understanding of what's happening and what we can do, but humans maintain ownership of what we should do and under what guardrails.
How AI Boosts Service Quality
We’ve seen the biggest gains in automation, speed, and service quality — about 90% of IT requests in our “Now on Now” environment are handled autonomously, reducing cycle times and freeing engineers for higher-value work.
AI also improves visibility by surfacing patterns across workflows.
Why AI Fails Without Governance And Organizational Context
AI underdelivers when deployed in silos. Without a unifying layer, organizations lack visibility into performance, compliance, and real impact. Isolated initiatives, especially those not connected end-to-end across the business, rarely create meaningful value.
AI also struggles without sufficient organizational context. To be effective, it needs to understand how work happens across systems, workflows, and data. A unified, single-pane-of-glass approach provides that visibility, ensures compliance, and ultimately unlocks AI’s full potential.
It also underdelivers without proper governance. Organizations must prioritize governance early in their AI journeys. You can’t layer it in afterward, so build it in from day one, ideally through an AI control tower to ensure consistency, compliance, and performance.
Why AI Without Workflows Is Just Expensive Advice
Every CTO should understand that AI without workflow and governance is just expensive advice.
The real value doesn’t come from adding a model on the side; it comes from embedding AI into the workflows, data, and controls that run your business or mission, and treating it as a governed platform capability from day one.
Starting with point solutions leads to fragmentation, security and compliance gaps, and limited impact.
How An Agentic Blueprint Allows For Autonomous Execution
An AI-powered engineering workflow built on an agent orchestrator can coordinate multiple agents to complete complex end-to-end tasks. It should follow a "sense, decide, act, secure" framework: agents first sense, ingesting data and context; then decide, applying models and logic; and act, executing tasks across systems.
This agentic blueprint embeds security and governance throughout the process, ensuring compliance, visibility, and control at every step, allowing teams to move from manual, fragmented work to coordinated, autonomous execution.
How Knowledge And Data Quality Impact AI Output
Never underestimate the importance of the data foundation, as AI’s efficacy depends directly on the quality of the data it uses. Without strong data quality and visibility, outcomes are inconsistent.
Never underestimate the importance of the data foundation, as AI's efficacy depends directly on the quality of the data it uses. Without strong data quality and visibility, outcomes are inconsistent.
AI agents produce unreliable results when knowledge is fragmented or outdated, as they depend on accurate, accessible data. Centralizing and governing knowledge — while using AI to surface gaps and stale content — creates a stronger foundation for consistent, high-quality outcomes.
This is especially true in the public sector, where knowledge often resides across legacy systems, institutional expertise, and incomplete documentation. In an AI-driven environment, agents are only as effective as the knowledge they can access.
I address this by embedding AI within a unified workflow data fabric, enabling agents to tap into trusted, up-to-date information and deliver more accurate, consistent results.
Why Real Impact Requires An Integrated AI Strategy
So, my advice to technical leaders is to focus on building an operating system for the agentic enterprise by embedding AI into core operations rather than bolting it on. This requires a unified platform approach across data, security, and identity, integrating governance from day one.
Isolated AI efforts won’t scale; real impact comes from an end-to-end, integrated strategy.
This also means leaning into strong ecosystem partnerships and creating a unified front door — such as MoveWorks — to make interacting with AI more intuitive.
When deployed holistically, organizations can reduce risk, maintain trust, and scale value across the enterprise.
Follow Along
You can follow Jonathan's work on LinkedIn.
More expert interviews to come on The CTO Club!
