IT trends 2026 part 2: GenAI’s real use cases
How AI is being used by UK organisations
In the second part of our coverage of IT trends in 2026, we look at the uses and impact of GenAI.
A debate rages over whether GenAI will improve or diminish developers’ productivity over the long term, but one thing is certain: software development is one niche where AI has caught on in a big way.
In part, this is because producing software has always been a continuous process of increased automation. It’s something that developers take in their stride, and GenAI is not that different to previous waves. Second, development is a structured process, with standardised processes that are easily reproduced and numerous repetitive tasks that a bot can perform pretty well and much more quickly in many cases. Third, based on their vast training knowledge, LLMs can suggest viable new approaches that developers might never discover on their own.
But the number one use case is to help with an essential activity that few developers enjoy: documentation. In a survey of 138 UK IT leaders, 43% said they or their development teams use GenAI to automate this unloved and frequently neglected aspect of software development.
Generating initial frameworks via prompt (34%) was another popular activity, while 32% mentioned using bots like a pair programmer for routine work. Additional common use cases included checking test coverage and generating tests and automating DevOps processes.
Base 138 UK IT leaders
AI’s impact on jobs
Many developers appreciate GenAI, but there are downsides. With AI starting to take on activities that would have traditionally been handed to junior developers, it is no surprise that 19% said that AI will mean fewer junior devs hired this year, compared with 4% who said they’d be hiring more starters.
Looking more widely, almost a quarter (23%) said AI will change the structure of the workforce in their organisation in 2026, with those respondents indicating that back-office roles in particular will see a reduction due to advances in the technology. However, 34% believe that new roles will be created over the period, with customer facing and sales-type roles likely to see a slight boost.
See part 1 of this research: Is AI a bubble? IT leaders have their say
Several respondents mentioned the rise of prompt engineering, or more broadly, new roles designed to get the most value out of AI investments.
“We are recruiting more technical staff to write and support AI-enabled tools, both experienced and every-level developers, plus some upskilling and reskilling of existing staff to cover AI tasks,” said a strategist in a government ministry. “However, we also seek to reduce headcount in areas where AI and automation can save human time, though the focus is more on augmentation than full replacement.”
“There’ll be new roles around AI governance and management, but fewer roles doing routine tasks which can be removed by automation,” echoed a CIO at a university.
“Sales, marketing and dev roles will increase, but there might be decrease in back-office roles,” offered a CTO at an IT service provider.
Reducing repetitive tasks
Eliminating mundane, repetitive or manual tasks that a machine could do more effectively, more safely or more cheaply has always been a major driving force for innovation, and thus far GenAI is no different.
Most such use cases reported by respondents can be placed into five main buckets: Discovery, Customer service, Software development, Summarising and reviewing, and Drafting.
Discovery. Generative AI is useful for unearthing hidden information and connections and producing intelligence for marketing or other operations by quickly sifting through masses of documentation. “We use it for gathering routine audit evidence,” said a senior manager at a consultancy, while an IT director at a technology company used GenAI to discover new ways to solve technical problems.
Customer service. Routine customer queries are commonly handled by chatbots these days, which can free up staff to handle more technical or complex questions. The head of IS at a nature charity was one of many to mention this use case.
Software development. GenAI has been rapidly integrated into development environments, and software production is perhaps the area where it has had most impact so far (see above).
Summarising and reviewing. Perhaps the biggest no-brainer for most organisations. GenAI tools are now very adept at voice recognition, translating, checking documents and summarising long texts into digestible abstracts and bullet points. “Minuting meetings and summarising texts and reports are just a few use cases,” reported a CIO in higher education.
Drafting. From simple emails to long-form documents to marketing collateral and PowerPoint presentations, GenAI can take on much of the boilerplate, leaving the author time to finesse the final version.
Other GenAI use cases, which didn’t come up in this research, include cybersecurity (which is more about traditional machine learning) and image creation.
AI agents
Nine percent of respondents said they were rolling out AI agents with a further 19% working on it.
These respondents reported a range of use cases for these agents, as illustrated in the table below, but most could be regarded as extensions of those mentioned above.
Within the industry agents are rather loosely defined, but the word implies a degree of autonomy in decision making.
A CTO in technology said: “What an agent does is to ask, ‘what is the user actually trying to do?’ Then let's apply some intelligence around what they are actually trying to achieve before I answer the question.”
Regarding implementation, a head of digital services in a tech firm noted that agents are likely to be successful when individual tasks are kept simple. “The payback is in the ease and speed of automating deterministic or procedural tasks, so the benefit is in implementation rather than the AI’s ability to be flexible to variable inputs.”
In the next article in this series we’ll be looking at which LLMs organisations are using, and which AI use cases they think are the most promising.