Most enterprise IT service desks run on a mix of human agents and RPA bots handling password resets, account provisioning, license moves, software installs, and other ticket types. The reality is that RPA bots break as soon as an application UI changes, forcing a development team to rebuild the bot from scratch. Meanwhile, the backlog of tickets keeps growing. The root cause is that the automation stack assumes a stable world while the IT environment never stops changing.
Why RPA breaks in service desks
Traditional RPA tools (UiPath, Automation Anywhere, Power Automate) bind to specific selectors, xpaths, and object IDs. When IT departments upgrade a portal, change a field label, or adjust a workflow, those selectors become invalid. The bot halts and the service desk team must open a ticket to fix it. Industry research shows that a majority of RPA implementations spend more than half of their budget on maintenance and rework, not on new automations. In service desks, where portals and forms change frequently, the rebuild-on-change cost is especially high. Bots that were stable for months can suddenly stop working after a patch or a UI refresh, creating a chronic interruption in ticket processing.
What changes with computer use agents
Computer use agents are different. Instead of relying on brittle selectors, they see the screen and act like a human: move the mouse, click, type, and read the result. This gives them several important advantages in IT service desk automation.
- They survive UI changes
- They require no brittle selectors
- They recover from exceptions and unexpected states
- They follow the SOP as written
- They work on legacy apps, Citrix, and virtual desktops
The one line a VP of automation should remember: computer use agents see the screen and adapt, while RPA bots break when the screen changes.
How to move without the risk
You do not have to rip out all RPA at once. A pragmatic path for IT service desks is to pick one high-pain process that is dominated by changing UIs and exception handling, for example, a ticket type that spans multiple portals and systems. Document the process in plain English. Then run a pilot with a computer use agent to see how it performs. Measure error rates, resolution time, and maintenance effort. If the pilot succeeds, expand to similar ticket types. Keep the stable, high-volume RPA automations in place where they fit best. The goal is to shift the long tail of variable work to agents while preserving the efficiency of the bots that are already working. This phased approach lets you reduce backlog and maintenance burden without a full-scale migration.
If you want to stop rebuilding bots every time IT changes a UI and start automating with agents that can see and adapt, the next step is to see it in action. Book a demo with the Coasty team at https://cal.com/coasty/15min.
Want to see this in action?
View Case Studies