No, if it is scoped correctly. A well-built agent removes the repetitive third of a role, not the role itself.

What happens to the freed-up time

In every engagement we have run, the hours went back into work that needed judgement: the awkward customer, the exception the rules do not cover, the account worth a phone call. None of our clients reduced headcount as a result of a build.

Where it does go wrong

Automation fails when it is sold as a headcount argument from the start. If the goal is to cut people, the scope gets pushed past what the technology can reliably do, the agent starts making calls it should not, and someone spends their week cleaning up after it. That is a worse job than the one it replaced.