Artificial intelligence is no longer confined to technology teams. It is becoming part of job descriptions across finance, marketing, human resources, management, engineering and operations.
Indeed’s 2026 Mid-Year UK Jobs and Hiring Trends Report found that AI was mentioned in a record 9.4% of UK job postings. Jobseeker searches for AI roles had risen sevenfold since the launch of ChatGPT. This happened while overall UK job postings fell by 11% between the start of 2026 and 17 July and stood 32% below their pre-pandemic baseline.
The contrast is striking. Employers are advertising fewer jobs overall, but a much larger share of those jobs now refers to AI. That signals a structural change in the skills businesses expect – not simply the growth of a new specialist occupation.
The challenge is deciding what AI capability genuinely means for each role.
AI literacy is not one skill
Some organisations have responded to AI by adding broad requirements such as “AI experience” or “proficiency with AI tools” to job descriptions. These phrases sound current but often reveal little about the work.
An engineer using machine learning to detect equipment failure needs a different level of capability from an HR adviser using generative AI to produce a first draft of a policy. A finance professional may need to interrogate automated forecasts and identify unreliable outputs. A manager may need to understand governance, confidentiality and when human judgement must override a recommendation.
Employers should define capability through tasks and decisions. Which process will the person improve? What information will they handle? What output must they be able to verify? What would unsafe or inappropriate use look like?
This produces a clearer brief and avoids rejecting strong candidates because they have not used a particular branded platform. Tools will change. The ability to frame a problem, assess an output and apply relevant professional judgement is more durable.
Avoid creating impossible job descriptions
As AI becomes fashionable, there is a risk of combining several disciplines into one vacancy: data analysis, automation, software development, strategy, change management and the original professional role. The result is a specification few people can meet and a salary that may not reflect the breadth of expertise requested.
Employers should separate what the person must know on arrival from what can be learned in the role. If AI is supporting an existing function, deep knowledge of that function may be more valuable than advanced technical credentials. A capable management accountant who understands controls and can learn a new analysis tool may outperform an AI specialist who does not understand the consequences of a flawed financial output.
This distinction also widens the talent pool. Research on skills-based hiring suggests that requirements for formal degrees have been falling in some AI-related roles while specific capabilities command greater value. Practical assessments, work samples and evidence of applied learning can therefore be more informative than a list of qualifications.
Recruit for judgement as well as confidence
AI-fluent candidates may present themselves impressively, but confidence with a tool is not the same as competence. Interviews should explore how candidates validate outputs, manage uncertainty and respond when the technology is wrong.
A role-specific exercise can help. Candidates might review an AI-generated report, identify unsupported conclusions and explain what additional evidence they would seek. This tests professional judgement without requiring an organisation to ban or celebrate AI use indiscriminately.
Employers should also be transparent about their own maturity. Advertising an “AI-first” environment without clear tools, data access or leadership support creates disappointment. Candidates need to understand whether they are joining an established programme, an early experiment or a role expected to build the approach from the beginning.
Build a capability plan not an AI shopping list
The rise to 9.4% of postings shows that AI is becoming an everyday workforce issue. The organisations that benefit most will not necessarily recruit the greatest number of AI specialists. They will define where the technology helps, develop existing people and hire targeted expertise where it adds something the organisation cannot build quickly enough.
That requires collaboration between operational leaders, HR, technology and recruitment. The question is not simply, “Do we need AI skills?” It is, “Which decisions and processes do we need people to perform better, and what combination of domain knowledge, technology and judgement will achieve that?”
Workforce Professional and Workforce Tech help employers translate changing capability needs into realistic briefs and evidence-based recruitment. Speak to our specialists about finding and developing the skills your business will actually use.





