How AI agents will transform enterprise IT operations

With Gartner’s latest forecast revealing enterprise spending on generative AI (GenAI) will reach US$644 billion in 2025, the technology’s initial promise of productivity gains is giving way to more nuanced questions.

As GenAI becomes more embedded in workflows, organisations are starting to figure out the different factors that determine its true impact. Is it the complexity of the task, the experience of the worker, or the strategic value of the work itself? Is the role of AI simply to optimise tasks, or can it evolve to drive deeper business value?

This is where AI agents come in. Unlike today’s AI assistants like ChatGPT or CoPilot, which largely respond to prompts, AI agents are designed to act independently of human intervention. They use large language models (LLMs) and machine learning (ML) algorithms to perceive, make decisions, take actions and achieve goals in their digital or physical environments.

Gartner predicts AI agents will be implemented in 60% of all IT operations tools by 2028. This critical distinction from AI assistants offers a glimpse of how AI agents might operate as a true enterprise collaborator rather than just a tool.

Infrastructure and operations (I&O) leaders can’t afford to overlook the potential business opportunities this shift presents. But implementing AI agents requires careful consideration of where the technology fits, what risks it introduces and how it can be deployed responsibly across IT operations.

Transforming IT operations

AI agents are poised to radically disrupt the IT operations status quo, due to their shift from simply augmenting specific tasks to executing complex incident handling without human intervention, freeing up operations personnel to focus on higher value activities.

For example, when given a prompt, LLM-based agents can plan using so-called ‘chain of thought’ reasoning to make decisions based on current data and past experiences, and through a process of trial and error.

AI agents also leverage ML algorithms to detect anomalies. Recent data from Gartner shows that AI agents will reduce the time it takes to exploit account exposures by 50% by 2027, helping IT teams respond faster to emerging cybersecurity threats.

In complex IT environments, the real advantage of AI agents comes from their ability to understand context and objectives. This allows them to orchestrate multiple functions simultaneously, creating composable IT operations capabilities that are adaptive, modular and scalable. By interacting with external tools and data sources via APIs, AI agents enhance their functionality, expand their knowledge base and drive collaboration across systems.

I&O leaders should start piloting AI agents to explore their capabilities, define their limits, and assess how they interact with existing IT systems and infrastructure. Understanding this is key to scaling AI agents across the enterprise.

Improving decision-making

AI agents could also help better understand how information flows in IT operations, leading to improved future decision-making transparency.

AI agents can break down high-level goals into smaller actionable steps. They can analyse dependencies between those steps to identify the correct order of execution and flag any potential conflicts. This allows them to generate structured tasks and even build out entire project plans without human intervention.

One potential example of leveraging AI agent frameworks in strategic decision-making is proactive resource allocation. AI agents can simulate budget constraints against future workloads and execute scenarios with feedback loops to help find the optimal balance.

With AI agents able to continually monitor various data sources such as incident records, logs and scrums, they can also assist with managing technical debt by prioritising potential remediation activities.

I&O leaders should start by developing a roadmap with clear milestones and policies that link the business impact of AI agent adoption with decision-making.

Disrupting the vendor ecosystem

AI agents will have a significant impact on IT operation vendors because it will push them to rethink how they design and deliver products and services.

This is a welcome change in an area where many tools remain tightly bound to user interfaces and siloed data. Despite efforts to integrate them, the result has often been a mix of APIs that are difficult to maintain, slow to evolve and vulnerable to change. Even AI IT operations (AIOps) have struggled to deliver on their promise, often limited by scale, complexity and speed for enterprise environments.

I&O leaders should encourage suppliers to develop AI agent versions of their capabilities that expose functionality at multiple levels and can be orchestrated by emerging AI agent platforms.

It’s also important to evaluate AI agent strategies by key vendors to determine whether they remain strategically aligned with the organisation’s AI roadmap. Finally, keep a close watch on evolving AI agent ecosystems and assess how they may reshape vendor relevance in the years ahead.

Building skills development

The design of AI agents to support IT operations will require the development of skills and the implementation of tools for the integration and management of increasingly intelligent and autonomous systems.

The bad news is that the journey will require an increasing knowledge of both GenAI and AI agent capabilities. The good news is that evolving, low-code tools and offerings are being designed to reduce some of the needed skills depth being offered by technology providers.

Beyond just acquiring the knowledge to develop AI agent capabilities, IT operations must ensure any implementations address potential risks, due to the ability of AI agents to act independently, while adapting to their environments.

To achieve business value from AI agents, invest in developing AI technical literacy and collaborate with other organisations that may be developing AI agent applications to design and implement the necessary governance and control measures.

*Cameron Haight is a VP analyst at Gartner focused on I&O automation and AI. He will be presenting on these trends at the Gartner IT Infrastructure, Operations & Cloud Strategies Conference, to be held on 13–14 May in Sydney.

Image credit: iStock.com/GamePH