As businesses strive to deliver AI-driven growth and efficiency, the technology itself is set to undergo a surge of innovation.
Just as the initial wave of predictive AI applications opened up the way for today’s generative AI capabilities, so agentic AI is set to transform customer experience and user productivity workflows with hitherto unseen levels of automation.
So why is this technology so powerful? Agentic AI acts on the user’s behalf to perform tasks with minimal intervention. It breaks down complex objectives into a series of simple steps, working directly with applications, databases, APIs and even other agents to complete their assignments.
With time, agentic AI may support the evolution of a digital workforce, where persistent agents embedded into workflows learn from their user interactions, and operate dynamically to handle exceptions.
As Nicholas D. Evans put it in a recent CIO.com article: “Instead of just throwing up a user interface and asking questions of a gen AI chatbot, CIOs are looking to use agentic AI to execute tasks and orchestrate workflows going deep into enterprise processes, such as CRM, supply chain, enterprise resource planning, HR, finance and more.”[1]
The arrival of agentic AI can be seen as a paradigm shift, where the keyboard and mouse-driven user experience of the last four decades gives way to more efficient and intuitive workflows, based on natural language processing and systems that anticipate the next step.
A new era
In a recent blog, IDC Program Vice President Eric Newmark, suggests that developments in agentic AI over the next three to four years will see software evolve so that “agent-driven interfaces will become more dominant, and reliance on traditional interface and UI design will begin to fade.”[2]
Agents are already handling research and document analysis for lawyers and accountants, employee questions for HR teams, and taking over simple, standard case types in customer support. EY is using agents in its third-party risk management service to deliver vendor reports in minutes rather than days.[3]
To take advantage, organisations will need a robust data foundation and a modern tech stack, capable of supporting AI agents, and delivering a consistent platform in which they can operate.
Think about AI PCs
Agentic AI will place new demands on the devices IT teams roll out. While much of the AI processing will still happen on the cloud, AI PCs are set to play a major role.
Agentic interfaces could accelerate productivity, making operating systems and applications easier to use. Microsoft has been trialling agents in the Windows settings menu that translate user prompts into system-level actions, while its Modern Context Protocol enables agents to work directly with Windows apps.
Meanwhile, AMD has demonstrated agents operating complex 3D modelling software using text-based prompts. This kind of interaction will be widespread in future versions of Windows and Windows applications, demanding more of the NPU, GPU and CPU resources found in modern AI PCs.
What’s more, not all agents and processes will be suitable for the cloud. Customer privacy, security, regulatory needs and performance requirements may require employees to run agentic AI on-device. Here, agentic AI based on efficient small language models (SLMs) may provide business-specific capabilities that run more effectively on NPUs, while agentic AI based on LLMs adopt a hybrid model, running partly on local hardware, partly in the cloud.
Agentic AI is a technology that will fundamentally redefine the way we work. Investments in AI PCs now, as enterprises refresh corporate fleets, could help support the full range of use cases in the future, on top of the other benefits that come with modern hardware.
The AI future demands a new type of PC. Learn more about AMD PRO Processors for Copilot+ PCs.
[1] The agentic AI reset is here, CIO.com, June 2025
[2] IDC Blog, The Agentic Evolution of Enterprise Applications, IDC, April 2025
[3] Agentic AI: 9 promising use cases for business, CIO.com, June 2025