How can you get hired for AI work?
ps AI certifications are not what’s getting people hired
Most weeks I’m here testing AI tools, running comparisons, and giving you an honest verdict on what is actually worth your time. This week we’re going a different direction.
If you or someone you know is trying to break into AI work or pivot into it from another field, this one is for you.
The certificate completers are everywhere right now. Coursera badges, Google AI certificates, BCS credentials, Microsoft Azure foundations. LinkedIn profiles stacked with them. And for what? To get screened out in round one anyway.
Here is the thing nobody selling AI courses wants to admit: UK employers are not hiring on the basis of what you know in theory. They are hiring on the basis of what you have already done.
What employers actually want
I have spoken to enough people building teams in 2025 and 2026 to notice a pattern.
The candidates who move forward are not the ones with the longest certificate section. They are the ones who turned up with something to show.
A workflow they built. A process they automated. A before-and-after that shows time saved or cost reduced. A project, a case study, a GitHub repo, something that demonstrates they have actually put AI to use in a real context and produced a result someone cared about.
A certificate proves you sat through the material. A portfolio proves you understood it well enough to do something with it. In competitive hiring, the second matters more.
The roles where this comes up most are exactly the ones growing fastest: automation, operations, marketing, analysis, implementation. None of those hiring managers are asking “do you have the Google AI certificate?” They are asking “can you show me something you built?”
Where certificates still have a place
This is not an argument against certificates. They still have value. Just not the value that the course-selling industry implies they have.
They are useful for building foundational literacy, especially if you are moving into AI from a field like marketing, operations, recruitment, or education. They give you the vocabulary, the context, the sense of where the edges of the technology are. That matters in interviews and client conversations, and it matters for avoiding decisions that come back to bite you.
They are useful for platform alignment.
If the job you want runs on Microsoft 365, Azure, or the Power Platform, a Microsoft AI credential sends a clear signal that you can contribute faster. That specific credential in that specific context has direct operational relevance, not just symbolic value.
They are useful at a career transition point. Not because the certificate alone gets you hired, but because it can get you through the first screening. A hiring system that filters on credentials is less likely to bin your application if you have something to point to. The certificate is not the decision, but it is sometimes the door.
And the one area where the value is genuinely growing: governance and responsible AI.
Ethics, risk, transparency, bias, legal implications.
These are not nice-to-have conversations anymore, especially for roles in finance, healthcare, legal, procurement, and any organisation that has had a bad experience with AI output.
If you are coming for those roles, a credential that shows you have thought seriously about safe adoption is more relevant than it was two years ago.
The professional body question
BCS is worth separating out from the generic certificate pile. It is the most UK-recognisable credential in the AI professional space, and it sits inside a broader professional framework that carries weight in the UK IT and business community. It is not the same as completing a Coursera course. Whether it is the right choice depends on what you are going for and who you are trying to impress, but if UK professional recognition matters to you, BCS is the clearest option in the current market. The BCS Foundation Certificate in Artificial Intelligence is the starting point, and they also offer a broader AI certification pathway including ethics and governance.
Vendor credentials have their place, but mainly when the employer already lives in that ecosystem. The Microsoft Azure AI Fundamentals is worth looking at if the role runs on Microsoft tools. Google’s AI Professional Certificate on Coursera is the clearest entry point on the Google side, and Google Skills has additional free learning paths worth exploring. AWS has the Certified AI Practitioner for a more technical route. A Google AI certificate means more at a company that runs everything on Google Cloud than it does somewhere else — same principle applies across all three.
One worth flagging separately: Anthropic launched its own free learning platform in 2026. Anthropic Academy now has 17 courses covering AI fluency, API development, Claude Code, and model context protocol, all free and all with certificates on completion. There is also a dedicated AI learning resources page on the Anthropic site with additional material. The Claude Certified Architect credential launched in March 2026 as the first formal technical certification for partners, with more planned. For anyone building directly on Claude or working in AI implementation, this is the most practically relevant free option in the current market.
The harder truth about what the market actually values
The candidates who are getting hired in AI-adjacent roles have a combination of things that no certificate covers. They understand how to translate a vague business problem into a workflow. They know what AI can and cannot do in context, not just in principle. They can manage the risk, explain the governance, and handle the stakeholder conversation. And they have a track record of using the technology in a setting where it actually mattered.
Those things come from doing the work. They do not come from completing a course.
In regulated industries, the gap is even wider. Finance, healthcare, law, public services: employers in those sectors do not just want someone who knows how the technology works. They want someone who understands the data handling, the compliance risk, the practical adoption constraints. A certificate is a weak proxy for all of that. Real sector experience is the signal.
What to do instead
If you are building toward AI work, spend less time collecting certificates and more time making things. Build a workflow and document what it does. Automate a process and measure the time saved. Build a prototype, write up a case study, put something in a portfolio that someone can actually look at and evaluate.
The certificate can come alongside that. It will mean more when it sits next to evidence of application.
In the UK, the hiring decisions in this space are going to a combination of things: practical experience, portfolio evidence, role-relevant knowledge, and the ability to communicate clearly about what you are doing and why it is useful. Those are skills you build through doing. The certificate confirms you know the theory. The work confirms you know what to do with it.
One is screening evidence. The other is hiring evidence. Know which one you need.
In short: AI certificates can open a door or add credibility, but they are not driving the hire. The candidates getting through are the ones who show up with work.