We’re currently hiring a QA role on my team and recently had a conversation with my manager about our candidates that left me scratching my head.
I interviewed a small pool of candidates for the a final round. Prior to the final round was a take home task and a first stage. I felt good about one who could handle QA responsibilities well. However my manager wants to pause keep interviewing and raise the bar significantly.
Some highlights from the Manager’s feedback:
They want this QA hire to ‘upgrade our AI workflow’ and active building / implementing of internal AI systems, well beyond using AI tools to write test cases or testing AI features.
The ‘AI can do it’ mindset In the past, losing a key person could cause serious problems. But AI agents have become so capable that they can handle almost anything. We’re in a position to be much more selective, even when it comes to top-tier candidates.
They not being sure of what upper management want from the additional QA role we’re recruiting for.
Also strict 3 days on site requirements driven by past trust issues with remote hires going missing.
My core questions:
Are you seeing management pushing AI implementation duties onto QA roles without adjusting compensation or title?
How are you countering the belief that ‘AI agents can handle almost anything’, so hiring standard QA capacity isn’t urgent?
How do you help management define what they actually want from QA before rejecting good candidates for failing to fulfill moving goalposts?
Would be great to hear how others navigate this mismatch in scope and expectations.
When I was working, I did not see any adjustment to compensation or title. There might have been a bonus component that I am not aware of. As an aside, I wonder if QAs are expected to just learn AI on their own time with little to no guidance. That is what I have experienced.
the hiring mistake is treating three jobs as one undefined quality bar. senior QA, automation or platform engineering, and internal AI systems each need different evidence, authority, and compensation.
i’d ask management to write the first 90-day outcomes before interviewing anyone else: which risks this person owns, what AI workflow must improve, what they can change, and how success is measured. if they cannot name that, candidates are being rejected against a moving fantasy role.
ai agents do not remove QA capacity. they shift the work toward defining oracles, reviewing generated changes, investigating failures, and deciding what evidence is trustworthy. hire for that actual gap, then make AI platform work an explicit separate responsibility if it is genuinely needed.
This is the real deal. If management believes this, then: Why we need testers? Why we need developers? Why we need team leads? etc. Let’s go directly from “C-levels” writing a prompt to Prod using AI in a fully automated way.
All the possible paths in-between are completely wrong and will not work. Like the omnipresent: “Adding human in the loop”…just nonsense.