And honestly? I think we should be talking about this more openly.
---
I have been working in QA for six years.
In that time, I have seen frameworks come and go. I adapted to Agile, DevOps, continuous delivery pipelines that compressed two-week release cycles into days. I have seen automation go from a "nice to have" to an absolute non-negotiable.
Every one of those changes mattered. Everyone required us to learn, adapt, and grow.
But nothing, and I want to be really clear about this, nothing has felt like what is happening right now with AI.
This is not another tool to add to your stack. This is a fundamental change in what is possible, what is expected, and what it means to be a great QA engineer.
---
Let's Honestly Go Back Three or Four Years
A new feature lands in the sprint. You get a Jira ticket, a couple of paragraphs, maybe an acceptance criterion or two. You read everything. You ask your developer questions. You dig into the codebase if you know it well enough. You think. You plan. You write scenarios.
That whole process, from reading the ticket to having a test plan you actually feel confident in, could easily eat up half your day. Not because QA engineers were slow. But because genuinely understanding a feature well enough to test it thoroughly requires context. And context takes time to build.
Your test coverage was directly limited by your knowledge. You could only test what you understood, and you could only understand what you had time to learn. Backend systems were black boxes unless a developer walked you through them. This was not a failure of QA engineers. It was just the reality. Knowledge was the bottleneck. And knowledge took time.
---
What Has Actually Changed
The knowledge gap is gone.
You can now sit with an AI, share a design document or a pull request or unfamiliar code, and have it explain the architecture to you. Not a surface-level summary, a real explanation. How components relate, what the failure modes are, what assumptions are baked in. In minutes.
For the first time, a thorough understanding of a system is not gated behind years of experience with that specific codebase. It is a conversation away. That is a genuinely new thing.
The pace of shipping has also changed. Teams now ship multiple times a week. Features that once took a quarter to build take weeks. For a long time, QA was where that speed ran into friction. AI has changed the math. Tasks that took hours (drafting a test strategy, planning API coverage, identifying regression risk) now take minutes.
---
A Moment That Changed How I Think About This
We had a new feature coming into Connect AI, something significant, with a detailed design document written by developers and architects. A lot of it was unfamiliar to me. New concepts, architectural decisions whose implications I did not immediately understand.
In the past, I would have built my test plan around what I could confidently grasp, which for an unfamiliar design would have been incomplete.
Instead, I had a conversation with Claude. I asked it to explain the architecture. I asked how components would interact, what would happen when things went wrong, what assumptions in the design could become failure modes.
By the end of that session, I had come up with testing scenarios I genuinely would not have found otherwise. The AI explained the architecture in a way that made the edges visible: the boundary conditions, the integration risks, the places where an implicit assumption needed verification.
I was asking questions about things I did not know I did not know. And that is the most important sentence in this whole piece. Testing has always been about finding what others miss. AI, when used well, expands what you are capable of seeing.
---
What This Looks Like in Practice
Our team has been building custom Skills, reusable AI workflows that encode expertise into repeatable processes. One that directly affects QA is a skill called create-test-plan.
You point it at a Jira ticket or a code branch. It reads the requirements, analyses what changed, and produces a structured test plan: Gherkin scenarios, edge cases, error states, a regression risk map, and automation candidates. Once you approve it, it creates a Jira sub-task automatically.
It is not perfect, and it is not meant to be. But what would have taken hours is now the foundation of a ten-minute review. The QA engineer's time shifts toward judgment: evaluating the plan, catching what AI missed, adding context only a human on the team would have.
---
But Let's Be Honest About the Limits
AI does not carry accountability. It does not know your team's history with a particular customer, or the instinct you develop from shipping and breaking things over years. A tool can only go so far. Human judgment is still the point. AI is what sharpens and accelerates it.
Our role is not disappearing. It is shifting.
We're moving from writing test plans to evaluating and sharpening the ones AI drafts. From hunting down behavior manually to verifying what AI surfaces. From knowledge being the bottleneck to judgment being the actual value we bring. The work is still ours. It just looks different now.
The QA engineer who uses AI well will not be replaced. They will be significantly more capable than the one who does not.
---
Get on the Train
In six years of this career, I have not felt this strongly about something. So let me just say it plainly.
The engineers who get curious, who experiment, who learn where AI falls short and where it excels? They are going to be operating at a level that simply was not available to anyone in QA before this.
The pace of software development is not slowing down. The complexity of what we test is not decreasing. The expectation that QA keeps up with both is only going to increase. AI is what makes that possible without burning out every engineer in the process.
You do not need to be an expert. You just need to start. Pick one part of your workflow and try it with AI assistance. See what it gets right. See where it falls short. Learn from both.
The shift is already happening. The question is not whether it is real. It is whether you will shape it or be shaped by it.
I know which one I am choosing. And I think it is one of the most exciting times to be in QA that I can imagine.