Agentic AI and GenAI are catalysts for transforming day-to-day IT workflows and entire operating models. The result: More efficient, responsive, and innovative IT.
Over the next 12 to 24 months, AI agents are expected to revolutionize every facet of business. According to PwC’s AI Agent Survey, 75% of respondents believe AI agents will reshape the workplace even more than the Internet did in the past. The changes are already materializing: 88% of respondents anticipate AI-related budget increases due to agentic AI in the next year, 79% say their businesses are adopting AI agents, and 66% report measurable value through increased productivity.
IT organizations in particular are on the frontlines of AI agent adoption. More than half (53%) of respondents to the PwC survey said their IT and cybersecurity functions were making use of the technology. This includes the use of AI agents to cut software development cycle times and to reduce production errors.
“The monotonous work in IT operations can be taken over by AI agents,” says Rani Radhakrishnan, Principal, Al and Analytics Managed Services Leader, PwC US. “AI agents can scale and do more without adding human heads. Human talent can then be upskilled to do higher value work and manage agents, including monitoring for hallucinations.”
Guidance for implementing agentic AI
As AI agents make their way into IT workflows as well as other areas of the business, CIOs are poised to play a wide-ranging role. IT executives are well suited to lead development of an agentic AI strategy, define architecture and the technology stack, and engage with key business stakeholders to manage AI agents as part of the broader workforce. CIOs should keep the following in mind for this latest agentic AI chapter.
IT operations is a good place to start. Many critical IT processes, including patching, cybersecurity updates, triaging support tickets, and software testing, are manual and take time away from more impactful work. AI agents are well suited to automate many of these mundane IT tasks, delivering significant benefits in short order.
For example, working with PwC, a retail company was able to leverage AI agents to automate many aspects of its software development pipeline, including requirements, code generation, testing, and workflow orchestration. The resulting AI agent platform helped shorten software development cycle times by 60% while cutting production errors in half. There is also the multiplier effect.
Take a holistic and strategic approach. It’s one thing for individuals to create bots and agents to streamline their own work or even a single department’s tasks. But the real gains come with effectively leveraging AI agents at an enterprise level across workflows, departments, and people. “This is about doing things differently with less labor hours in a completely different way,” says Cenk Ozdemir, CIO Advisory Leader for PwC Consulting. “You’re writing agents not just to solve tickets, but to predict them. That kind of thinking never existed previously.”
CIOs should be a lighthouse for IT architecture, determining how to more effectively approach an agent strategy and which platforms to invest in. Ozdemir cautions CIOs to learn from robotic process automation (RPA) adoption, which was limited by enterprise architecture complexity. “You have to think about what kind of AI agentic future you envision when building AI architecture today,” he explains.
Measure up. CIOs should develop new practices for measuring AI agents, evaluating their performance, and advancing their functionality. HR executives can be valuable partners, helping to coordinate strategies to upskill and educate the workforce as well as define strategy and tactics for AI agent-human interaction. “AI agents are an extension of humans. And even though they can be autonomous, they shouldoperate within bounds,” notes Dan Priest, PwC’s chief AI officer. “Accountability lies with humans.”
“A lot of companies are investing in Al capabilities, but the value is not realized immediately,” says Radhakrishnan. “Unless you invest time and energy into maintaining and optimizing the models, you’re not going to get the ROI.”
Embrace lifecycle management practices. Agent frameworks are essential for managing AI agent development and deployment and for reducing siloed efforts. For example, PwC’s Agent OS platform functions as an enterprise AI command center, seamlessly connecting AI agents into adaptive, business-ready workflows up to 10x faster than traditional methods. The platform integrates with core enterprise systems, regardless of cloud platform or framework, and delivers an extensive library of pre-built AI agents with flexible customization options. The integrated risk management and oversight frameworks support enhanced AI governance and compliance.
The bottom line
AI agents are here, and disruption is around the corner. CIOs should lead the agentic revolution, starting in IT operations and scaling impact across the enterprise.
To learn more, visit AI agents for IT: PwC
