Cast your mind back just a year, and you’ll realise how much of our daily lives has changed since generative AI first gained traction. It is reshaping industries at a scale comparable to the Industrial Revolution, fundamentally altering how we work and interact with technology, and it shows no signs of slowing down.
Imagine a future where AI doesn’t just follow instructions, but can learn, adapt, and act autonomously. This is the promise of agentic AI. Unlike generative AI, which requires prompts to proceed, agentic AI can teach itself and, in theory, undertake tasks it hasn’t been trained on. Think of them as intelligent modern-day assistants that can think for themselves and work independently. This represents a significant shift from current AI systems, such as chatbots, which typically need human input at every step. In Asia-Pacific (APAC), IDC estimates that about 70% of organisations expect agentic AI to disrupt business models by the end of 2026.
It’s clear that agentic AI represents the next stage of evolution and is already being used, or soon will be, across enterprises. However, it will require increased network bandwidth and deterministic connectivity, with compute taking place closer to users. These infrastructure elements are already being deployed.
Building a smarter ecosystem of IoT supported by agentic AI
For starters, the potential applications of agentic AI are seemingly endless. In an increasingly interconnected world, the Internet of Things (IoT) offers real-time data insights that can help redefine business models and drive industry transformation. Businesses are turning to emerging technologies built on IoT to boost time-efficiency, automate responses, and improve sustainability.
But for IoT to truly reach its potential, devices need more than just basic connectivity; they need to communicate and collaborate intelligently. As AI becomes more tightly woven into IoT systems, seamless machine-to-machine communication will be even more critical. Today, we can adjust the temperature in a room or car through a mobile app, without being physically present. Imagine how this might evolve with agentic AI: a car switching on its air conditioning every weekday morning, 10 minutes before the owner leaves, all by checking their online calendar, no prompts required.
Indeed, with smarter software tools, developers now have access to both historical and real-time analytics that offer a clear picture of what’s happening, deliver just-in-time recommendations, identify faults or anomalies, and help drive action. Devices won’t just react to real-time data; they’ll work together proactively to improve how cities function, such as:
- Traffic signals adjusting themselves to ease congestion.
- Optimising energy usage across neighbourhoods.
- Anticipating emergencies before they occur.
The potential for a more efficient, people-friendly urban environment becomes a tangible reality, built on intelligent coordination between connected devices.
In a world where IoT devices come from countless manufacturers and rely on different protocols, interoperability remains a major hurdle. But agentic AI, together with flexible, high-performing networks, has the potential to bridge those gaps, enabling smooth communication across devices regardless of brand, platform, or ecosystem.
Agentic AI powering networks of the future
Traditionally, networks have relied on human intervention to monitor performance, fix issues, and make adjustments. But with agentic AI, networks could begin to run themselves — learning from data, adapting to changes, and making decisions without the need for manual input. For example, AI-enabled networks can analyse traffic in real time, detect bottlenecks, and automatically reroute data or reallocate resources to keep services running smoothly.
Just as important is agentic AI’s ability to catch problems before they escalate. That kind of predictive capability means higher reliability and less downtime. The shift from reactive to proactive management could make networks more resilient, more efficient, and ultimately more responsive to what users need.
Unlike traditional systems, agentic AI isn’t tied to one location. It moves across different systems and endpoints, constantly exchanging data. This decentralised design makes the underlying network infrastructure even more essential, enabling fast, reliable communication between AI agents, edge devices, and the cloud.
The road to embracing agentic AI
From predictive maintenance and real-time analytics to customer support and network planning, AI is no longer a glimpse of the future, it’s becoming embedded in the present. The rise of agentic AI brings the promise of even more intelligent decision-making. While the past few years have been defined by investment and experimentation, we’re now entering a phase where consumers are starting to feel the impact through smarter services, more intuitive interfaces, and more efficient outcomes.
This shift is also opening the door to disruption. As AI reshapes consumer behaviour and expectations, new players — unburdened by legacy systems — have a real opportunity to enter the market and redefine it. In this environment, agility and innovation will matter more than scale. The businesses that succeed will be those that understand how to meet evolving needs with AI-driven experiences.
Network traffic has been growing steadily year after year, and AI is expected to accelerate that trend. To realise the full potential of agentic AI, a strong network foundation is essential. This becomes even more important as machine-to-machine communication expands across industries, grows in complexity, and introduces new possibilities for 5G use cases and monetisation. Service providers will need to prioritise not just compute, but also enhanced connectivity. The future isn’t just about faster speeds or more devices; it’s about using AI to enrich how we live, work, and connect with one another.
