Claire Agutter on shadow IT and reinventing the IT team

In Conversation

I love shadow IT

Claire Agutter on AI reality, vendor complexity, and why the IT department needs to reinvent itself

⁠AI tools are no longer arriving in organizations through IT procurement. They are arriving through people. Marketing teams running their own automations. Finance teams building their own workflows. Operations teams connecting tools without raising a ticket. What used to sit at the edges of IT governance is rapidly becoming the mainstream, and the question for IT leaders is no longer how to stop it but how to make it work.

⁠Few people are better placed to answer that question than Claire Agutter. Director at Scopism, the global home of Service Integration and Management, Agutter has spent more than two decades helping organizations navigate exactly this kind of complexity. 

⁠She started on the helpdesk, built one of the first ITSM e-learning companies, authored some of the most widely used service management publications in the world, and has grown the Scopism SIAM community to more than 4,500 practitioners across 120 countries. She has watched every wave of transformation the industry has been through, and she has a habit of saying the things others are still working up to.

⁠We spoke with her about where AI adoption actually stands, what organizations are getting wrong, and why the service desk's best days might still be ahead of it.

⁠There's a lot of noise right now about AI transforming IT. You've been around long enough to have seen every previous wave of transformation. What's genuinely different about AI this time and what's the uncomfortable truth most organizations still won't say out loud?

⁠My observation about AI adoption is how it's turned your typical IT service or product conversation on its head. Typical IT thinking is rooted in this idea of: we've got a challenge or a business problem or business need, how do we then develop a product or service that will address that need? With AI, you've got customers specifically requesting to use this technology, but without necessarily having the use cases in mind. We're coming at things from the other angle, starting with the technology and then looking at how it can be applied within organisations. The uncomfortable truth, however, is that many organisations are still tinkering around the edges and doing very basic automation activities using AI. What isn't happening is the transformation piece, the fundamental redesign and redevelopment of business processes aligned with the technology. 

⁠Vendor sprawl, shadow IT, unmanaged spend all keep growing despite better tooling and governance frameworks. When you're working with an organization that's drowning in it, what's the first thing that actually needs to change? Is this fundamentally a governance problem, a culture problem, or something else entirely?

⁠Personally, I love shadow IT! I think we should embrace it. It shows the business engaging with technology, it highlights the fact that so much IT spend now happens outside the IT department, and we have to accept that as the future. We need to look at how we build an operating model that supports and enables ‘shadow IT’ to happen in a safe, secure, and resilient way. 

⁠Viewing shadow IT as a negative just buys into that whole ‘department of no’ thinking that then ironically encourages the spread of shadow IT. And really, this is where I spend all of my time now: in the service integration and management space. We need to accept that, as organisations, we have these complex supply networks. How do we build a resilient, flexible, scalable operating model that can manage that with ease? 

⁠Agentic AI can now detect, diagnose, and close tickets without any human in the loop. In your honest assessment, are most IT organizations actually ready for that level of autonomy? And what does being truly prepared actually mean in practice?

⁠Many of the IT organisations that I work with are struggling with IT service management foundations. They are not necessarily ready for full agentic AI and incident resolution without a human in the loop. The data that they have to underpin that, the decision-making processes, all of that needs to mature before they can take that next step. However, when I work with our industry partners who are the large managed service providers in the SIAM space, what we see from them is really advanced capabilities. This is where many organisations would benefit from outsourcing and looking to buy this type of capability rather than try to build it internally. There are organisations that have already spent millions of dollars developing these capabilities, and they are now looking for customers to work with. 

⁠SIAM has grown significantly as a discipline over the last decade, yet many organizations are still managing multiple vendors without a formal integration framework. What's the cost of that and where do you see SIAM heading next, especially in the context of AI?

⁠SIAM is growing significantly and continues to grow. In our online community for practitioners, we now have more than 4,500 people from 120+ countries. This demonstrates that SIAM addresses very real challenges that organisations are facing from the public sector to the private sector. The fact that there are still organisations out there struggling but not aware of the guidance that SIAM can provide for them shows that we always still have more work to do to raise awareness. In some cases, organisations feel that SIAM isn't for them. They're not big enough; it's too much of a commitment. We've recently been developing content that's addressed to those types of organisations, helping them to apply some SIAM principles without necessarily having to go down a full transformation route. 

⁠Seventy percent of MSPs report using agentic AI internally, but only 10% are using it in real IT workflows. What explains that gap and what does it tell you about where the industry actually is with adoption?

⁠There is still a gap between the stories that are being told around AI and reality. We carry out the Global SIAM Survey every year and in 2025 the responses showed that only 37% of organisations have incorporated AI into their SIAM model.  Of those respondents who were using AI, 85% were using it for incident management, 62% for performance analytics, and 54% for problem management and service request management. We're seeing initial AI adoption at the operational level, and that will expand over time to include more advanced processes and workflows . 

⁠2026 is being described as the compliance year for AI in IT - the point at which governance moves from best practice to formal audit requirement. For IT teams that have been focused on adoption, is this a rude awakening or something they should have anticipated?

⁠It's absolutely something that IT teams should have been expecting. Legislation does not drop from the sky with no notice or warning. However, where many organisations are not ready is in their understanding of compliance across their supply network. This is something we see from a SIAM perspective, where organisations don't fully understand their supply network, which then means they don't know which legislation applies. It's very challenging for them to be compliant. Aligning with SIAM best practices helps them to have that visibility and that transparency and understand what information is needed where. 

⁠AI is taking over Tier 1 support tasks at pace. What does that mean practically for the people in those roles? Is the skills shift that follows an upskilling story or a deskilling one?

⁠It's been more than two decades since my first help desk role, and I used to spend a significant part of every shift manually resetting passwords. It wasn't particularly fun, it wasn't particularly stimulating, and there was a delay in restoring service to the person who had to ring up to request that password reset. If AI can continue that process of removing unnecessary, repetitive, and quite frankly dull support tasks, then that is a good thing. However, the one concern that I do have is: where now will the entry points into an IT career be? Many people do not choose to work in IT or IT service management roles. It's something that they fall into, and for many people that route comes via help desk experience. It comes with an imperative to embrace the concept of lifelong learning. Everyone who works in IT now has to continually refresh their skills, has to have that curiosity about what's changing, and has to be ready to embrace the change. 

⁠As endpoint management becomes more automated and self-healing, what happens to the service desk's role and how should IT teams be repositioning themselves in response?

⁠There will always be a need for the human in the loop at some stage of the process. The number of full-time employees on the service desk might shrink, but the role will also mature. Service desk employees can focus more on pattern identification, trend analysis, and working out where perhaps additional guidance or changes can make life easier for end users. I've argued for a long time that the IT department as a whole needs to rethink its role and to look at itself as a service enabler and a service aggregator and a point of expertise that the business can come to. The earlier question addressed shadow IT, and this is a case in point. The business is going to spend money on technology. What the IT department can do is provide some governance and some guidance to make sure that that technology is scalable, integrates when necessary, is secure, and is resilient; all of the other requirements still need to be considered. 

⁠You've built several communities over your career. Looking back at what actually worked, what does building a genuine practitioner community require?

⁠The macro trend has been a shift from broad, non-specialised platforms back to communities of practice and smaller forums. Platforms like LinkedIn do not have the same quality of conversation that they had maybe five or ten years ago, and there is a lot more sponsored or advertisement content. People are now looking for curated spaces where they can engage with other practitioners in their field and have real conversations.

⁠ To build a genuine practitioner community, there has to be that element of curation. 

⁠Firstly, the community has to be established with guidance to ensure it's a safe space. Then it does need to be managed to encourage people to contribute, to listen to feedback, and to make improvements. Stephanie Ward, who is our community manager, does a wonderful job keeping our community vibrant and alive. There is also a light touch element of policing that needs to happen. For example, removing content that is spam or is just advertising something, as well as dealing with any posts that might breach our code of conduct. 

⁠What I've learned in my time in IT service management is that people are desperate to join communities. They're desperate to have these meaningful conversations, and it's wonderful that the platforms exist that allow us to do that with people from all over the world. 

⁠Miles from the office

⁠Most people treat the helpdesk as a starting point they moved away from. Claire Agutter treats it as the foundation everything else was built on.

⁠"It's fair to say that I would have never had a career in IT had it not been for my help desk experience," she says. The move from customer service and retail to an IT helpdesk was the turning point. What she found there was a set of problems that matched the way her mind worked - identifying outcomes, solving issues, understanding the gap between what technology promised and what people actually experienced when it failed. "I never forget that behind every IT system is an original business need," she says. "What are we actually trying to do?" she asks. It sounds simple. In practice, she suggests, it is the question most organizations forget to answer.

⁠Her path through IT was not linear. She moved roles, changed environments, accumulated experience across operations, consultancy, and training. She has always been drawn to new challenges and new networks, and at some point the logical next step became starting something of her own. Her first company, IT Training Zone, was one of the earliest ITSM e-learning businesses, founded around 20 years ago. Scopism came later; it started not as a grand plan but as what she calls a sandpit, a place to test ideas. Some of those ideas worked. Others, she is candid enough to admit, really didn't. "I've learned a lot of lessons along the way," she says. "Some of them have been extremely difficult." 

⁠Away from the office, she keeps things deliberately grounded. When she steps away, she goes outside. Walking, nature, time with her husband, her dog, her friends. In the last two years she has pushed that further than most people would consider reasonable - 40 miles in 24 hours, then 35 miles the following year. For someone whose professional life is built around frameworks, governance structures, and operating models, there is something refreshingly unstructured about disappearing into the countryside for a day and a night with a target and a pair of boots.

⁠It is not the only arena where she has had to push through. She has been the only woman in the room enough times that she has stopped being surprised by it. "I do laugh to myself when I join a meeting with 16 people and I'm the only woman there," she says. Her advice to women earlier in their careers is practical rather than inspirational: get involved with industry bodies, go to events, build a network. That is what opened doors for her, and she believes it still does. But she is clear that it is not only about individual advancement. "Some of that is holding out a hand to the other women in tech that we meet," she says, "and creating opportunities for them too."

⁠What she wants, in the end, is not complicated. She loves the work. She loves seeing new people come into the industry. And she wants the community and publications she has built to be there for them when they do.