The CIO Case for Moving from RPA to Computer Use Agents
Your finance team’s reimbursement bot works for weeks, then stops when the SAP upgrade shifts a button two pixels right. Your compliance process halts the second a field is missing, and the developer who built it is on vacation. IT is drowning in change tickets. Every time an application updates, the bot breaks. Every time an exception occurs, the process halts. These are not isolated incidents. They are the hidden cost of legacy RPA and manual SOPs at scale.
Why RPA breaks here
Traditional RPA relies on selectors, XPath, and object IDs to locate controls. When an app updates, the selector list becomes stale. The bot clicks the wrong location or misses the target entirely. Industry surveys show that 60 to 70 percent of RPA projects exceed their original timelines because of maintenance. When a field changes or a UI layout shifts, a developer must rebuild the bot from scratch. The rebuild takes time, introduces new bugs, and pushes out the next release. The cost compounds with every change. Citrix and virtualized desktop environments make this worse. RPA struggles to locate controls in those environments, so teams must build workarounds or skip processes altogether. The result is a backlog of low-value, high-friction tasks that only humans can run.
What changes with computer use agents
- ●survives UI changes without rebuilding
- ●no brittle selectors to maintain
- ●recovers from exceptions instead of halting
- ●follows the SOP as written
- ●works on legacy and Citrix desktops
Selectors lock you into a snapshot; seeing the screen lets you adapt.
Selectors vs. seeing the screen
RPA binds to a specific element on a specific screen at a specific time. Computer use agents see the screen and act like a human: move the mouse, click, type, and read the result. When the UI changes, the agent notices the new location and adjusts automatically. It does not need a developer to rebuild the bot. This flexibility extends to exception handling. If a field is missing, the agent can retry, ask for clarification, or log the issue and continue. It does not halt the process. This approach also removes the need to maintain long selector lists. The agent works as long as the goal is clear. It can follow a standard operating procedure written in plain English, turning text into action without building a separate flowchart bot. Legacy apps and Citrix environments become accessible because the agent interacts with the visual interface, not with hidden identifiers.
Computer use agents turn brittle, manual SOPs into durable automation.
How to move without the risk
Start with a single high-pain process that has frequent UI changes or many exceptions. Map the process to a clear SOP, then run a pilot with a computer use agent. Measure the time saved, error reduction, and the number of maintenance tickets that disappear. Use those results to justify a broader rollout. Do not discard RPA. It still fits high-volume, stable, backend tasks that do not change often. Use computer use agents for the long tail of changing UIs, exception-heavy workflows, and processes that rely on SOPs. This phased approach lets you test the new model in a controlled environment, learn from real-world usage, and scale only when confidence is high. The goal is to reduce the maintenance backlog and free capacity for higher-value work, not to replace every existing bot overnight.
If your RPA bots break on every UI update and your SOPs live in spreadsheets and Word docs, it is time to consider a more durable automation approach. Computer use agents survive change, recover from exceptions, and follow SOPs without building brittle flows. Book a demo with the Coasty team to see how they work on your environment at https://cal.com/coasty/15min .