Over the years, self-service, automation and chatbots have all been hyped as technologies that would supersede the service desk. None succeeded. AI is undoubtedly the most capable technology we’ve seen so far, and for the first time the predictions of the service desk’s demise don’t seem entirely far-fetched.
But is AI truly a replacement for the service desk, or simply the next stage in the evolution of service management?
There was a time when logging an issue with your company service desk meant picking up a desk phone and hoping someone in the basement would answer before your coffee went cold. At Revo, we’ve seen first-hand support evolve from people answering phones and manually logging requests into spreadsheets, to email-based ticketing systems, self-service portals, knowledge bases, automation, and now, AI-powered virtual agents.
While new technologies have redefined the role and remit of the service desk, none have come close to eliminating the need for it. Automation has taken over repetitive tasks, users have gained more ways to access support, and service desk professionals can now focus on more complex and valuable activities. But the need for human expertise and organisational knowledge remains as strong as ever.
Will AI change this?
In our view, not yet.
Without doubt, AI is delivering significant benefits to service desks, but there remains a considerable difference between an AI-enabled service desk and AI replacing the service desk altogether.
In this article, we’ll explore the five reasons why, despite the hype, human expertise remains the service desk’s most valuable asset.
1. Users rarely describe problems clearly or completely.
AI works best when given clear instructions. But, as any service desk agent will know all too well, users rarely describe problems in a structured or complete way. Consequently, support tickets arrive with descriptions such as:
- “It doesn’t work.”
- “The system is broken.”
- “My laptop is slow.”
- “I can’t access [platform].”
For an experienced analyst, this opening is the start of the investigation. They instinctively know what questions to ask next, how to interpret incomplete information, and how to spot hidden clues that the user may not even realise are important.
While AI is becoming increasingly capable of asking clarifying questions and guiding users through troubleshooting steps, uncovering the real problem through conversation and by applying real-world experience remains beyond its pay grade.
2. AI is only as effective as the quality of the data available to it.
AI provides responses based on the information available to it. We’ve all seen examples where an LLM confidently provides an answer that sounds plausible but turns out to be inaccurate or based on outdated information. The same principle applies to service management. If teams haven’t updated the knowledge articles, asset records are incomplete, or a workaround exists only in a team member’s head, even the most advanced AI will struggle to provide reliable answers.
These data deficiencies are common scenarios in most organisations. Experienced human analysts have the capability to compensate for them. For example, they might remember a similar issue from six months ago or recall a workaround devised during a previous major incident that never made it into the knowledge base.
In contrast, an AI model trained on public information cannot magically understand a case management system developed in-house ten years ago or an undocumented integration between two systems. So, until organisations capture and structure every piece of organisational knowledge in an AI-accessible way, experienced service desk professionals will need to bridge data gaps.
3. Human empathy and communication are still critical to a service journey.
A panicked user who has lost a presentation before an important meeting. A critical business process that has failed at the worst possible moment. When people are stressed, frustrated, annoyed or worried, a technically correct response delivered with no understanding of the user’s situation can make a bad situation even worse.
While AI is becoming increasingly able to recognise sentiment and tailor its responses accordingly, it still struggles to read between the lines and adapt an approach in real time in the way experienced service desk professionals can. Effective support requires understanding the impact an issue is having on the individual and the wider business and recognising when the situation calls for more support than a standard troubleshooting protocol offers.
4. Some issues require physical observation and intervention.
Despite the advancements in technology, sometimes the best way to troubleshoot a problem is still to walk over to that person’s desk.
For example, a disconnected cable or a monitor connected to the wrong input. While AI-powered advances in remote support are reducing the need for on-site intervention, there are some situations where seeing the problem first-hand provides insight that no chatbot or automated workflow can replicate. Especially when the real problem is something entirely different to the issue a user initially reports.
5. The more AI we adopt, the more governance we need.
For the last few years, the conversation around AI has largely focused on its ability to answer questions, generate content and automate routine tasks. The agentic phase of AI is different.
Agentic AI systems are now emerging that can make decisions, execute actions, interact with business applications, trigger workflows and collaborate with other autonomous agents with increasingly limited human intervention.
At first glance, these advances seem to reduce the need for human involvement, but in reality the more autonomy organisations give to AI, the greater the need for governance. If an autonomous agent wrongly approves access to a business-critical application or makes a decision based on incomplete information, organisations still need to understand who is accountable and what controls the agent followed.
In this new agentic era, service management will become vital in providing the guardrails that will allow autonomous agents to operate safely and effectively. So, although the role of the service desk may change, the need for service management is unlikely to disappear.
The most successful organisations are taking a pragmatic approach to AI adoption
Rather than attempting to de-humanise every interaction, the organisations achieving the greatest success with AI are targeting it where it can deliver the greatest impact. Ticket triage, request routing, self-service, routine administration – these are areas where automation and AI can reduce response times and free up valuable service desk resources.
At the same time, they recognise that some activities continue to benefit from human involvement. Complex troubleshooting, stakeholder management, business-critical incidents, user reassurance and situations requiring judgement or contextual understanding are all areas where experienced service desk professionals remain invaluable.
This is reflected in the direction of the ITSM market itself. Platforms such as HaloITSM and Xurrent have embraced AI through capabilities such as ticket categorisation, knowledge discovery, virtual agents, summarisation and intelligent automation. But neither vendor is positioning AI as a replacement for service management. Instead, they are embedding AI into ITSM processes to make support teams more effective.
Finding the right balance between AI and human expertise
If you’re exploring how AI can enhance your service desk, the first question shouldn’t be “What can we automate?” but “Where can AI create the most value?”
At Revo, we help organisations identify these opportunities and implement solutions that complement and enhance their existing service management practices.
To discuss how AI could support your ITSM strategy, get in touch with the Revo team.
