Where AI fits — and where it doesn’t

The past year has been a turning point for CIOs. Artificial intelligence has moved from experimentation to executive priority. But in the rush to “adopt AI,” I’ve seen many organizations stumble — not because of the technology itself, but because they skipped the fundamentals.

After three decades in enterprise IT — supporting global clients from Silicon Valley to India — I’ve learned a simple truth: AI delivers value only when your foundation is strong. Whether you’re managing legacy datacenters, cloud infrastructure or modern DevOps environments, AI cannot replace IT discipline. It amplifies it. That’s why I often tell peers: AI on top of chaos just automates the chaos faster.

To make AI work, CIOs must start by understanding their existing systems, workflows and business outcomes. Only then can we identify where AI genuinely fits — and where it doesn’t.

AI is a legacy debt detector, not a silver bullet

Every enterprise has legacy DNA. Even the most modern digital-first companies run on layers of infrastructure, policies and human processes that have evolved over decades. When organizations try to drop AI into this ecosystem without cleanup, it often exposes what was already broken: outdated workflows, siloed data and undocumented dependencies.

In my own practice, I’ve seen AI projects succeed only when teams first stabilize identity access management (IAM), clean up their data stores and establish clear service ownership.

The promise: AI in service operations (ITSM)

In one area, the AI value is immediate and measurable: service operations. We’ve moved beyond simple chatbots; sophisticated AI agents are now automating entire L1 processes, driving significant productivity gains. For example, some enterprises have successfully reported up to a 50% reduction in Tier 1 IT support workload by deploying intelligent assistants that resolve issues like password resets and access requests autonomously.

However, this immediate success can be deceptive. A smooth AI deployment in service management is only possible because the inputs (tickets, knowledge articles) are already structured and contained.

The reality: AI in DevOps and broader business functions

Move to the complexity of the DevOps pipeline or a finance workflow, and the challenges multiply. Here, AI must interact with unstructured code, complex business logic and constantly changing regulations.

I’ve observed that the most successful AI initiatives in DevOps are focused on augmentation, not full replacement. For instance, in the software development lifecycle, AI is being strategically deployed to review code, with research showing it can help improve developer productivity and code quality by automating error detection and flagging potential security flaws. AI is great at repetitive tasks like test case generation and code suggestions, but human oversight remains critical for architectural decisions and security audits. Just like in ITSM, AI’s impact is proportional to the maturity of the underlying process.

The lesson: Not all AI value is immediate. Some domains require slower, more deliberate progress to protect trust.

3 principles for responsible AI leadership

The role of a CIO is evolving from infrastructure custodian to business strategist. AI makes this transition both exciting and daunting. To lead responsibly, I recommend three guiding principles:

• Start small, prove fast: Pilot AI within structured, data-rich environments like ITSM or DevOps before scaling.
• Measure outcomes, not hype: Track metrics that matter—cost savings, resolution times, employee satisfaction.
• Govern before you grow: Build guardrails for data privacy, ethical use and vendor accountability before deploying enterprise-wide. This includes navigating the complex requirements for cross-border data transfers using official mechanisms like the standard contractual clauses (SCCs).

In my view, AI’s greatest contribution won’t be replacing humans — it will be amplifying human intent. When applied with discipline, AI can help us achieve what IT was always meant to deliver: reliability, responsiveness and resilience.

A powerful lever

AI adoption is no longer optional. But how we adopt it defines whether it becomes a strategic advantage or an operational burden. For organizations with 2,000+ employees, AI represents a powerful lever to boost workforce productivity — especially in IT support and service operations. Yet, the CIO’s mission remains the same as ever: Align technology to business value, not buzzwords. As we step into this new era, remember: AI will redefine IT, but not by replacing it — by reminding us that technology’s purpose is to amplify human intent, not override it.

This article is published as part of the Foundry Expert Contributor Network.
Want to join?