By Krishna Sai, CTO at SolarWinds
Two-thirds of people use artificial intelligence (AI) regularly, while an overwhelming majority believe the use of this game-changing technology will result in a wide range of benefits in the future. But despite this apparent vote of confidence, more than half still view AI with more than a hint of suspicion.
These are just some of the insights contained in a landmark global study by KPMG, published last year, that helps shed some light on the quickly emerging age of intelligence. And in particular, they reflect perfectly the inherent tensions that exist between benefits and risks.
For avid readers on this subject, all too often it can feel as if the negative stories and opinions outweigh the positive narratives. And yet, real-world tangible evidence about the benefits of AI is starting to emerge – if you know where to look. Take IT service management (ITSM), for example.
Often overlooked, helpdesks sit at the operational heart of financial institutions, keeping trading platforms running, securing access to systems and resolving issues that, left unchecked, can quickly translate into financial and regulatory risk.
Why IT service desks matter more than you think
Managing helpdesks – and ensuring a swift response and resolution to problems – is time-consuming and labour-intensive. Traditionally, the only ways to improve response are to increase the number of people you have handling enquiries, allocate more budget or both. But AI is changing that. In particular, generative AI (GenAI) is speeding up the provision of helpdesk support, enabling IT teams to shave hours off every incident they handle.
The details of this surge in productivity are detailed in the latest 2025 SolarWinds State of ITSM report, which tracked more than 60,000 anonymised data points from over 2,000 service desks worldwide to provide the clearest evidence yet of the measurable ‘AI effect’ on IT support performance.
For example, it found incidents took around 27.5 hours to resolve before the introduction of GenAI. After GenAI was enabled, that figure dropped to slightly over 22.5 hours, saving almost five hours on every single ticket.
Among the top 10 GenAI adopters cited in the report, the improvement was even more marked, with average resolution times slashed from a little under 51 hours to just over 23 hours per incident. That’s a huge time saving of nearly 28 hours on every ticket they handled.
GenAI cuts resolution time
The research also looked at the efficiency gap between those who have adopted GenAI and those yet to make that investment. Again, the findings couldn’t be clearer. It found non-GenAI customers resolved enquiries in slightly less than 33 hours, while GenAI-enabled teams were closing incidents in 22.5 hours – a saving of almost 10 hours per ticket.
It doesn’t matter whether you’re part of an IT team responsible for delivering tech support or a user looking for help – the outcome is pretty much the same. The introduction of GenAI has a significant impact in terms of speed of response and ticket resolution.
Imagine the difference such savings would make to any organisation operating in the finance sector, where time is literally money.
But there’s an important caveat in the data that’s easy to miss. The report makes it clear the biggest gains aren’t made by those organisations simply bolting on AI and hoping for the best.
Why AI only works when the basics are in place
The real winners are those organisations that already have strong processes in place – including automation, self-service and knowledge management – and are using GenAI to amplify what’s already working. In other words, this is a story less about AI tools and more about how work is designed around them. And it’s an idea that is increasingly gaining traction.
A recent Insights2Action article published by Deloitte put it this way: ‘The problem for most organisations is that while they are making investments in AI, they aren’t making commensurate investments in their people and work design’.
It went on: ‘Companies that prioritise human-machine collaboration through redesigned roles, processes, and operating models are significantly more likely to realise measurable returns on AI compared to those taking a technology-first approach’.
The lesson from this is a simple one. If you work in financial services and are responsible for ITSM, AI is no longer optional. But to extract the maximum value, it’s important to have the right systems, people and processes already in place.
And even if ITSM isn’t your day-to-day concern, there’s still value to be had from this study. After all, service desks are among the first operational functions to apply AI at scale. They also operate under intense pressure. The lessons being learnt in such a high-pressure environment could apply equally to other functions within an organisation and to other sectors.
That’s because in reality, most organisations are still learning what AI means in practice. And if that learning means looking elsewhere for help, advice and lessons learnt – including in unexpected places such as IT support – then it’s well worth taking note.
