Back from MWC 2026: Agentic AI at the Edge Driving the IQ Era

Why the IQ Era and Agentic AI Matter for the Intelligent Edge?

By Semir Haddad, Chief Product Officer, MicroEJ

I just returned from Mobile World Congress Barcelona 2026 and came away with a conviction that this year’s event marked a structural inflection point for our industry. The official theme, The IQ Era, reflects a shift in narrative that is both deep and systemic. Intelligence is no longer a layer that sits loosely on systems. It is becoming embedded, pervasive, and autonomous across devices, connectivity, and infrastructure.

Unlike past eras where innovation was defined by faster processors or new form factors, the IQ Era emphasizes the integration of continuous intelligence into every layer of the technology stack. That intelligence is increasingly agentic, able to observe, decide, act, and adapt in context with minimal human intervention. Industry thought leaders at MWC were blunt about this: agentic AI will “eat the world” in 2026, redefining how software and systems operate in the real world.

What is Agentic AI?

Agentic AI refers to AI systems that can autonomously perceive context, make decisions, and take actions to achieve defined goals with minimal human intervention. This is no incremental evolution. It is a foundational shift from reactive computation to intent-driven autonomy.

True agentic systems become co-actors in workflows, resolving decisions that historically fell to humans.

Concretely, that means systems that:

• Interpret context and plan action sequences
• Act autonomously toward goals without being explicitly triggered
• Adapt their behavior over time based on outcomes

At MWC, keynotes and demos across the floor echoed this clearly: intelligence is being pushed into physical infrastructure, mobile ecosystems, and edge devices alike.

What Agentic AI Means for the Edge?

For embedded developers, this is both an opportunity and a wake-up call. Embedded has long operated under real constraints: limited compute, fragmented ecosystems, tight hardware coupling, slow integration cycles. That legacy made embedded a laggard compared to cloud and mobile. But agentic AI reframes the problem entirely.

When intelligence moves to the edge, devices stop being passive sensors and start acting as context-aware agents deciding locally, optimizing in real time, coordinating across distributed environments. That unlocks capabilities that were simply out of reach before:

• Real-time autonomy without constant cloud dependence
• Edge participation in multi-agent workflows
• Proactive responses across industrial, consumer, and enterprise contexts

At MWC, sessions on AI 4 Enterprise and AI Nexus drove this point home: embedded is no longer a bottleneck. It is becoming the platform.

Pictured: Sarah McMurray, Audio product marketing lead at Qualcomm and Semir Haddad, Chief Product and Strategy Officer at MicroEJ.

AI as the New User Interface

A clear proof point of this evolution is our collaboration with Qualcomm.

On the Snapdragon Sound platform, we demonstrated a generative AI assistant running directly on the device, without a smartphone, enabled by direct-to-cloud connectivity and secure downloadable applications through Snapdragon Sound Apps powered by MicroEJ VEE Wear.

The intelligence was embedded, not proxied. AI effectively became the user interface.

This example illustrates how edge devices can transition from fixed-function hardware to software-defined, agentic platforms that evolve over time.

The same demo was shown earlier this year at CES in Las Vegas. You can watch the full CES demonstration here.

The Limits

There are real constraints that still shape how agentic AI can be implemented:

• Resource limitations on low-power devices
• Security, governance and predictable behavior
• Complexity in updating and validating autonomous logic
• Fragmentation across silicon and embedded toolchains

Agentic AI introduces autonomy, but that autonomy must be trustworthy and verifiable. In many industrial or safety-critical contexts, it is not enough that a system can act autonomously. It must act consistently within defined boundaries.

The Opportunities

At the same time, the opportunity created by agentic AI is enormous. When intelligence becomes native to devices and infrastructure, embedded becomes a platform for continuous innovation, not a bottleneck. That opens doors to:

• New business models based on adaptive products
• Ongoing experience evolution through software platforms
• Real-time operational optimization across industries
• Distributed autonomy in manufacturing, mobility, health, logistics

What the industry presented at MWC was not isolated novelty. It was a signal that agentic AI and embedded intelligence are converging with connectivity and software platforms to create new classes of systems.

How Fast Can Organizations Adopt Agentic AI for the Edge?

The short answer is that adoption is a sprint, not a marathon. Organizations that embrace agentic architectures now will be exponentially more innovative. Those that wait risk falling behind as competitors shift to adaptive and autonomous systems.

But speed of adoption depends on your foundation. Fragmented stacks and tight hardware coupling slow everything down. What accelerates implementation is platform thinking: abstracted execution environments, modular AI runtimes, and secure foundations that treat software as a first-class, evolving asset.

MicroEJ’s Vision for the IQ Era

At MicroEJ our mission is to help embedded systems thrive in the IQ Era. We believe in sustainable software abundance: a world in which software innovation can be delivered reliably, securely and at scale across diverse hardware.

Our solutions provide the foundation for agentic AI at the edge:

• MICROEJ VEE (Virtual Execution Environment) that abstracts hardware and enables consistent behavior across platforms
• Unified semantic model that simplifies development across languages and domains
• Hermetic security and reliability with sandboxing and deterministic resource management

These technologies help teams move from concept to agentic capability rapidly. They reduce integration friction, improve security posture, and allow intelligence to live where it matters most: close to data and action.

Closing Thoughts

MWC 2026 was different. It was not about incremental upgrades. It was about an industry recognizing that intelligence is fundamental, not optional. Agentic AI will shape product experiences, network infrastructure, and operational systems throughout this decade. It is a defining theme of the IQ Era.

For embedded systems, the challenge is not simply to keep pace, but to reframe how devices participate as autonomous agents in distributed systems.

At MicroEJ we believe that a solid Agentic AI foundation is critical. We are committed to enabling this transition so that embedded systems become not a constraint, but a driver of intelligent innovation in the IQ Era.

The age of fixed-function devices is ending. The age of software-defined devices built for the edge-native era is beginning.