Artificial Intelligence
It’s Time to Start Designing for the Invisible Power User
Agents are everywhere and organizations have to start designing with them in mind.
Sydney Scott
Editorial Strategist, AI
Workday
Nearly 50% of all internet traffic now comes from machines instead of people. This automated activity is growing eight times faster than human browsing. But AI agents don't click buttons or scroll through pages. Instead, they navigate software through code and data.
This shift is driving a fundamental redesign of business software known as agent experience (AX). As AI agents begin to navigate complex systems like Workday independently, design must evolve from human-centric to hybrid-centric. We are moving beyond merely adding AI features to building entire ecosystems designed for both human and machine users.
Nearly 50% of all internet traffic now comes from machines instead of people.
Report
C-Suite Global AI Indicator Report: AI Is the Ultimate Level-Up
Why Businesses Must Design for Agents Now
Machine requests are skyrocketing, with agentic AI traffic growing by a staggering 7,851% in 2025. While this surge began on search pages, it is now reaching transactional checkout pages. This marks a fundamental shift: Agents are joining the workforce.
To support these digital teammates, organizations must pivot to agent-aware design. Without it, systems face performance crashes and security gaps as machines interact at a scale humans cannot match. However, with robust AX, these agents can handle high-speed heavy lifting—from payroll audits to complex procurement—freeing humans for higher-level strategy.
We are already seeing this in practice. Workday uses an "Agent System of Record," allowing agents from various partners to collaborate on hiring through a shared machine language. Similarly, a European automaker reduced costs by 50% using "agent squads" to refactor legacy code in parallel. This level of speed was only possible because they stripped away the UI fluff that confuses machines, replacing it with the structured, predictable pathways that define successful AX.
How AX Changes the Way We Work
Designing for AX shifts the human role from performing manual data entry to managing high-speed system results. When systems are agent-ready, they handle the repetitive coordination that usually clogs a human's day.
This changes the workplace into an environment where humans do not just use tools, but instead oversee a digital workforce that can execute complex, multi-step transactions in seconds. This shift requires humans to move above the loop—providing ethical judgment and strategic vision while the digital workforce handles the paperwork.
As AX becomes the standard, roles evolve or emerge specifically to manage this machine-readable environment. For example, in an AX-enabled office, a hiring manager no longer coordinates between IT, payroll, and security; instead, they supervise an agent factory where specialized agents handle the entire onboarding process instantly.
This decoupling of growth from labor costs means a small team of three people can now oversee tons of specialized agents. This environment forces a shift from an expertise culture to a learning culture, where the most valuable human skill is the ability to direct and refine machine-driven results.
Value is created by removing the UI fluff that confuses machines and replacing it with structured, predictable pathways.
Building the Machine-Readable Office
Creating a user experience for agents is similar to creating an accessible user experience. Both rely on a clear layout that communicates meaning without needing to "see" a screen. Using semantic HTML and Accessible Rich Internet Applications (ARIA) attributes helps both people and AI agents find their way through a page's hierarchy.
For some organizations, the infrastructure for this shift is not yet fully in place. While standard APIs exist, they often lack the semantic richness agents need to act without human help. Moving to a machine-readable office is a significant challenge that requires moving beyond basic front-end patterns. It is not a simple plug-and-play task. It requires a new technical stack that includes several key pieces:
Standardized protocols: Adopting frameworks like the Agent-to-Agent (A2A) protocol to allow different agents to talk to each other across various apps.
Semantic interfaces: Moving from visual-first design to machine comprehension, where the team binds every UI element to backend logic through data attributes.
Resilient architecture: Building distributed caching, load balancing, and circuit breakers to absorb the unpredictable surges in traffic that occur when agents work in parallel.
- New security layers: Replacing simple "click to buy" buttons with cryptographically signed digital contracts, known as mandates, to prove a human actually authorized a machine's purchase.
Toward Agent-Aware Design
Of course, designing for the agent experience doesn't come without risks. Agents don't naturally know their own limits. They can get too much access if designers do not provide strict boundaries and clear permissions.
With this in mind, some organizations are using accountability by design, where they save every choice an agent makes as a clear data record. Other companies are tapping agents to supervise agents, watching the main agent for any mistakes. New tools like zero-knowledge proofs help agents prove they have permission to act without sharing any private data.
Lessons from building AI tools are already helping some teams create strong, machine-ready systems. And rules like A2A protocol and model context protocol (MCP) are breaking down the walls between agents.
But ultimately, the move to a human-AI workforce is a total revolution that requires companies to move beyond a human-only user experience to an agent-aware one. Teams that fail to design for agents face performance crashes, security gaps, and a total loss of speed in a market moving at machine-ready rates.
As blended teams emerge, leaders who build for both a human and AI experience will win the most in this new economy.
Report
C-Suite Global AI Indicator Report: AI Is the Ultimate Level-Up
AI has been a gamechanger across nearly every industry and marketplace. But realizing the true value of AI requires buy-in between different stakeholders. In this AI Masterclass, learn how you can make the most of AI for finance, HR, and more.