Date Posted: Thursday, 24th September 2026
Technology roles are changing faster than the job titles used to describe them. An organisation might know it needs stronger AI capability or someone who can integrate new tools into existing systems, and still struggle to say what that person should be called or which experience actually matters. Search for an exact job title and you can easily conclude the talent isn't out there, when in emerging technology the right person often exists under a different one entirely.
That's a narrower problem than technology hiring generally: we've written elsewhere about building an AI-ready team and about planning the technology talent your growth depends on. This piece is about the vocabulary gap specifically, for employers whose real question has shifted from "who has done this exact job before?" to "who has the foundations to do this work well?"
Our wider guide to the future of work and emerging technology roles looks at how roles are developing across AI, platforms, data, FinOps and automation.
Skills pressure in UK technology is well documented. Skills England estimates that 30 priority digital and technology occupations could require around 488,000 workers between 2025 and 2035, though that figure doesn't mean 488,000 new jobs: roughly 239,000 comes from potential employment growth in its central scenario, with around 249,000 additional workers needed to replace people leaving these occupations altogether.
Skills England is candid about the uncertainty in those figures, cautioning that AI makes long-term occupational forecasting harder than it used to be. Even so, the broader direction holds: the UK will need substantial technology capability, and the nature of that capability keeps shifting.
AI illustrates the point well. Research commissioned by the Department for Science, Innovation and Technology found that 35% of organisations surveyed struggled to fill AI roles, with lack of work experience cited as a barrier by 31%, almost level with insufficient technical skills at 30%. Knowing how to use an AI tool isn't the same as knowing how to apply it where security, reliability, regulation, cost and commercial outcomes are all in play, and the real gap in the market is people who can demonstrate that, not simply people who understand the technology in the abstract.
Established roles come with established expectations. Recruit a network engineer, business analyst or project manager, and there's a reasonably mature, shared understanding of the profession, even allowing for variation between individual roles.
Emerging technology is messier: the work can develop before the market has agreed what to call the person doing it. Skills England has flagged this directly, noting in its 2026 assessment that conventional occupational classifications struggle with highly specialised and emerging digital roles.
AI makes the problem visible. One company's AI Engineer might primarily build models. Another's might integrate third-party models into products and workflows. A third might be looking for someone focused on deployment, monitoring and reliability. Same title, three different jobs. Defining an emerging technology vacancy by title alone can needlessly shrink the talent pool available to you.
New technology roles rarely need an entirely new type of person. More often, they're new combinations of established skills.
LinkedIn's UK Jobs on the Rise research bears this out. People moving into AI Engineer roles commonly came from Software Engineer, Data Scientist or Machine Learning Engineer positions; people moving into Head of AI roles came from software engineering, data science and business strategy. That should change how employers approach emerging talent. Instead of asking who is already an AI Integration Specialist, it's more useful to ask:
That question points towards software engineers, integration engineers and machine learning engineers whose current job title would never have surfaced in the original search. The same logic holds elsewhere: someone moving into AI assurance might come from model validation, technology risk, audit, testing or data governance; platform engineering talent often sits among experienced cloud, DevOps and site reliability engineers; FinOps capability spans cloud engineering, technology finance and infrastructure cost management. The role is new. The foundations mostly aren't.
Most discussion about AI and employment focuses on which jobs will disappear, but for many technology teams the more immediate change is happening at task level, not job level.
Skills England identifies a shift in AI-enabled digital work away from routine coding and testing and towards oversight, verification and communication, alongside growing requirements around governance, audit trails, bias testing, explainability, orchestration and change management. PwC's 2026 Global AI Jobs Barometer points the same way internationally, distinguishing between roles where AI amplifies existing expertise and roles where it makes a task accessible to people with less specialist knowledge. PwC found stronger job and salary growth in the former group, along with a significant wage premium attached to AI skills.
For hiring, that points to something more specific than adding "AI" to every job description. As execution gets easier, the skills surrounding the technology become the harder thing to find. Worth testing for in an interview:
None of this shows up in keyword matching. All of it matters more than it used to.
When hiring for an unfamiliar role, it's tempting to look at what other companies are recruiting for and reproduce their job descriptions. That approach quickly produces a specification listing every desirable skill associated with the technology, rather than the ones that actually matter for your business. A more useful starting point is what the person needs to make happen, broken into four questions.
1. Work. What does this person actually need to deliver, improve, integrate, govern or protect? Be specific: "drive AI adoption" tells a candidate almost nothing, while "integrate AI capabilities into customer-service workflows while ensuring outputs can be monitored, evaluated and safely escalated" gives you something you can translate into capabilities.
2. Foundations. Which skills must already be strong, because they'd be difficult, slow or risky to develop after appointment? An AI integration role might demand solid software engineering and systems integration foundations, for instance, while familiarity with one particular model or vendor is usually learnable on the job.
3. Adjacency. Where else could someone have built comparable capability? This is where employers can genuinely widen the talent pool. If the work needs reliability engineering, API integration and observability, candidates don't need to have held the exact emerging title before; look at what they've actually done.
4. Evidence. What would prove the candidate has the capability? That could be a detailed walk-through of a previous project, a representative work sample, or a structured scenario based on the real problems the role will involve. The point is to assess evidence, not vocabulary.
Emerging technology is particularly prone to inflated job descriptions and CVs. Tools change quickly and terminology changes faster still, so a candidate's familiarity with the latest vocabulary tells you little about how they'll perform when something actually breaks.
Interviews should dig into specific examples rather than general claims. Useful questions include:
For technical roles, a representative work sample can help too, where it's proportionate to the vacancy. The goal isn't an unnecessarily long technical examination; it's giving candidates a chance to show the kind of thinking the job actually requires. This matters most when assessing people from an adjacent profession, whose CV may lack the exact terminology in your job specification even though their underlying experience is entirely relevant.
Recruitment is only part of the answer. The DSIT-commissioned AI Labour Market Survey found that 88% of surveyed organisations were using on-the-job training to develop AI capability, and the proportion of AI hires coming through apprenticeships rose substantially between 2020 and 2025. That matters because emerging roles rarely come with a mature pipeline of people who've spent a decade doing exactly the same job elsewhere.
Sometimes recruiting externally is the right call. Sometimes it's moving someone internally who already understands the organisation and giving them the technical development they're missing. Often it's a combination of both. The decision that actually matters is working out which capabilities need to be bought in and which can realistically be built.
How do you write a job description for an emerging technology role? Start with the work rather than the title: what the person needs to deliver, which foundational skills must already be strong, and which adjacent professions might have produced someone with the right capability. Avoid copying job descriptions from other companies, which tends to produce an unfocused list of desirable skills rather than a clear brief.
Should you consider candidates without the exact job title? Yes. Many people moving into roles like AI Engineer or Head of AI come from adjacent professions such as software engineering, data science or business strategy. Filtering purely by current job title can screen out candidates with the relevant underlying experience.
Is it better to recruit externally or train existing staff for new technology roles? Both have a place. Recruitment brings in capability the organisation doesn't currently have; internal moves bring people who already understand the business and can be given the missing technical development. Most organisations use some combination of the two, and DSIT-commissioned research found 88% were already using on-the-job training to build AI capability specifically.
The technology workforce of the next few years won't be defined by a tidy set of new job titles. Some genuinely new specialisms will emerge, but established roles will also absorb new responsibilities, professions will overlap, and people will move into emerging areas from adjacent careers. That makes rigid title matching an increasingly risky way to hire.
The candidate you need may not have the title you expected, may not have used the same technology stack, and may have built the relevant capability in a different industry altogether. What matters is whether they have the foundations, judgement and evidence to do the work you actually need done, which means recruiters and hiring managers need to understand more than the vacancy title: how the work is changing, where transferable capability tends to sit, and which gaps can realistically be developed after appointment.
For a closer look at the roles taking shape across AI, platforms, data, FinOps, automation and technology delivery, read our guide to the future of work: emerging technology roles and where to find the talent.
And if you're hiring for a technology role that doesn't fit neatly into an established job title, TRIA can help you map the requirement, identify the adjacent talent pools, and take it to market.
As a founding member and Director of TRIA, Lara has been instrumental in shaping its strategic direction and ensuring its commitment to client success. Her expertise in the recruitment industry is matched by her dedication to fostering a collaborative and innovative work environment.
Connect on LinkedIn
