Computer-use agents in the enterprise

When they tempt

No API, vendor portal only, high manual load. The demo writes itself.

When they hurt

UI changes weekly, actions are irreversible, or audit needs structured evidence.

Prefer API integrations when available; use computer-use only with tight scopes and human confirmation on writes.

Risk classification

Read-only observation is different from submitting forms. Classify and require dual control for writes.

Record action sequences for audit. Screenshots alone are weak evidence without structured logs.

When to walk away

If the vendor offers an API, use it. Computer-use is a last resort, not a preference.

In practice

Map the workflow on a whiteboard before you open a framework: inputs, systems of record, humans, and irreversible writes. If that map is fuzzy, the agent will encode the fuzz.

Pick ten to fifty real historical cases as an eval set. Include the ugly ones. Run the agent offline against them until critical fields and hard rules are acceptable. Only then connect write tools.

Ship with a pause switch, a human queue, and a weekly review of override reasons. Promote repeated overrides into rules. That loop is how production systems improve—not another prompt brainstorm.

Common failure modes

  • Treating a demo on clean samples as readiness for production volume.
  • One shared service account with broad write access across systems.
  • No owner for the exception queue, so failures pile up as noise.
  • Changing prompts and models without regression gates on real cases.
  • Measuring only model latency or thumbs-up, not completed-case cost and audit completeness.

What good looks like after ninety days

The first workflow is boring: stable override rate, known failure modes, operators who trust the queue. Config changes go through review. Traces answer "what happened to this case?" without archaeology.

At that point you can add a second document type or a second agent role. Expanding before the first path is boring is how programs stall with five half-built pilots.

AI agent systems

常见问题

Are computer-use agents production-ready for core banking?

Treat them as high-risk. Start non-production or read-only observation, with strict dual control on actions.

Alternative?

Vendor APIs, file drops, RPA with known brittleness, or process change to reduce portal work.