The internet isn’t broken in a technical sense. It still connects the world and moves data at extraordinary speed. But for users, the experience has become cluttered and inefficient. Too much time is spent juggling tabs, passwords and logins instead of achieving outcomes. The result is fatigue and frustration, not productivity.
That’s about to change. Emerging natural language interfaces and artificial intelligence (AI) agents that can analyse data, connect systems through Application Programming Interfaces (APIs) and perform complex operations are changing how online interactions happen. This marks the beginning of the agentic internet, where intelligent agents act autonomously to achieve outcomes for people and businesses alike.
AI agents take the wheel
Until now, most AI tools have helped people complete tasks more efficiently. The next phase is about delegation, where agents act on users’ behalf. Instead of spending time comparing prices or filling in forms, people will set their preferences and let their digital agents negotiate, transact and optimise in the background.
This evolution will free up people to focus on intent and decision-making while agents handle the execution. Research shows that AI-powered consumers could drive up to 55% of spending by 2030, worth over $690 billion in the UK alone. Similar patterns will emerge across industries such as banking, insurance and healthcare, as digital interactions become agent-to-agent rather than human-to-machine.
The rise of agentic adoption
The move towards this new model has been building quietly for years and is now accelerating. Consumers who once hesitated to rely on algorithms are increasingly comfortable letting AI take on specific tasks. Major platform providers have embedded AI into their core infrastructure, putting powerful tools directly into the hands of users and developers.
Enterprise APIs, built over the past decade to connect systems, have unintentionally laid the groundwork for autonomous transactions. At the same time, connected devices are generating a constant stream of data that agents can act on without human involvement. The foundations for an agentic internet are already here – what’s changing is how quickly businesses must adapt to it.
From UX to AX
For years, businesses have focused on user experience (UX), making websites and apps easy for people to navigate. Increasingly, they will have to consider agent experience (AX). As autonomous AI agents begin to represent users, businesses will need to create digital environments that machines can read and understand clearly.
The goal of AX is to ensures that systems are intelligible, reliable and efficient for software intermediaries that will increasingly handle transactions, negotiations and exchanges. Traffic from autonomous agents is already growing, and organisations that ignore these new digital customers risk losing visibility in the marketplace.
Designing for agents in mind
The design principles that work for people do not work for agents. Human-centred design thrives on ambiguity, exploration and emotional engagement. Many consumers enjoy browsing, discovering new products or being influenced by storytelling. For people, the process can be as satisfying as the outcome.
Agents operate very differently. They value logic, precision and efficiency. When tasked with buying a product, an agent does not browse. It executes a clear query using defined parameters and evaluates multiple factors at once – price, delivery time, warranty, even sustainability impact – in milliseconds. For agents, success is measured by the best outcome, not by the journey.
This difference demands a new design mindset. While human interfaces can be flexible and interpretive, agent-facing systems must be structured and exact. Data needs to be consistent and self-explanatory. APIs must behave predictably. Services should be engineered from the ground up for machine understanding.
Operating in the agentic era
Agent experience goes beyond being a technical challenge. It changes how organisations operate as a whole. Product teams will need to treat APIs as products in their own right, while data teams will need to ensure clarity and consistency across every digital touchpoint. Compliance and risk specialists will also play a key role in maintaining transparency and accountability in how autonomous agents behave.
Testing will evolve too. Instead of watching users navigate websites, teams will study how AI agents interact with APIs and data. Sandboxes will let companies simulate agent behaviour, uncover weaknesses and see how their systems perform under real-world conditions.
To support this, businesses will need dedicated, cross-functional teams focused on AX. From an AX strategist to an API product manager and data governance specialist, new roles will emerge to embed AX principles into the development process. They’ll focus on keeping APIs reliable and data accurate, just as teams already do for security and performance.
Transformation by industry
Some sectors are already moving faster than others. Retail and media are natural early adopters. Retailers will move from selling individual products to fulfilling outcomes, such as a complete wardrobe or a week of meal plans. Meanwhile, media and tech companies will see agents managing subscriptions, bundling services and optimising access based on preferences.
Other sectors, including insurance and healthcare, will move more cautiously because of regulation and trust concerns. Even so, the potential is significant. Medical devices could coordinate care directly, adjusting medication or scheduling follow-ups automatically. Insurance products could become self-adjusting, responding to changes in behaviour or circumstance without human intervention.
Competing in an agentic marketplace
As digital agents begin to mediate most online activity, a company’s visibility will depend on being legible to machines, not just people. Metrics such as page views or conversions will fade in importance, replaced by how often agents choose to engage or transact with a brand.
Those adapting early will enjoy faster transactions, lower costs and stronger, automated relationships. The agentic internet is arriving fast, and readiness will determine who stays visible.
