AI-first IT service desk – new roles for people and machines

 

Artificial intelligence (AI) is not only transforming IT service desk processes, but entire job roles too. Firstly, human work is shifting 'upstream' towards higher-value analytical and creative tasks. Secondly, hybrid teams comprising humans and AI systems are emerging, where virtual agents handle initial contact while human specialists focus on complex, emotional, and high-risk cases. Understanding AI as a strategic lever that links automation, knowledge management, governance and new job profiles is crucial to the success of this transformation.

Where AI reduces human labour in the service desk, it primarily relieves human agents of repetitive, standardisable first-level support (L1) tasks, thereby reducing the number of simple interactions per agent. In well-designed use cases, virtual agents and chatbots can resolve 60–90 per cent of standard enquiries entirely via the self-service channel, meaning no ticket is created. This is cost-effective, as an enquiry to the bot generally costs less than half of what it would cost to contact a human agent.

Where AI reduces human support workload

Added to this is the automated classification, prioritisation and routing of incoming tickets based on their content, which largely eliminates the need for manual data entry, misallocations and subsequent corrections. At the same time, AI agents support human agents by providing relevant knowledge articles, suggested next steps and summaries of similar incidents, which shortens search times and reduces unnecessary escalations.

Human tasks remaining in the service desk

Although automation reduces the initial workload, human tasks are shifting towards operations, design and quality assurance of AI systems. Key tasks include AI operations and calibration. System prompts, dialogue flows and guardrails must be designed to ensure that bots act in a technically correct, consistent and brand-compliant manner. At the same time, confidence thresholds, handover points to humans and model adjustments are continuously optimised to prevent incorrect solutions and 'hallucinations'.

Requirements for background systems

For the AI-first service desk to operate effectively, knowledge management takes centre stage: content must be curated, structured, tagged with metadata and modelled as ‘atomic’ procedures that are easily consumable by AI and can be maintained in a version-controlled manner. In addition, the need for quality assurance and governance is growing, as AI responses are checked on a random basis, feedback loops are established and clear escalation rules are defined, whilst data protection, access control and explainability remain guaranteed.

 

 

Human tasks are becoming more complex

For our clients, the impact of the shift towards AI-powered service desks is clear. Tickets that reach human agents tend to be more complex and involve multiple systems. They are often business-critical and relate to ambiguous or emotionally charged situations, such as data loss or compliance breaches. Handling these issues requires in-depth technical diagnostic skills, an understanding of business implications, and strong communication skills. In other words, it requires precisely the kind of profile that is evolving from traditional first-level support (L1) towards specialised, context-rich incident handling.

Hybrid teams comprising humans and AI

As automation increases, the tasks and team structures are both changing. In complex cases that reach human agents, AI systems enhance knowledge and analysis by immediately displaying the most likely causes, suitable solutions, and relevant past incidents. This reduces the time taken for the first resolution. At the same time, intelligent routing ensures that tickets are directed to the correct priority group, reducing unnecessary escalations and waiting times, and indirectly improving first-contact resolution.

The traditional first-level role of human agents is evolving into that of 'AI-enhanced specialists', focusing on complex, emotional and high-risk cases. Experienced agents also act as knowledge and AI coaches, refining content, honing prompts and further developing automations based on practical experience. New roles are also emerging, such as AI Product Owner, AI Engineer, Knowledge Engineer and Data Analyst. These roles shape the AI operating model, are responsible for key performance indicators, and drive innovation in human-machine interaction.

AI-first service desk: more than just streamlining

In a mature AI service desk environment, it is possible to reach a level at which many standard scenarios are resolved automatically, while the first-resolution rate for human-handled tickets increases significantly. This requires the knowledge base and team design to be consistently geared towards integration and transparency.

Furthermore, AI acts as a catalyst for a new strategic direction within hybrid teams: bots provide speed, scalability and consistency, while human specialists contribute judgement, empathy, contextual knowledge and governance. IT managers who actively facilitate this shift in roles by investing in knowledge and AI capabilities and establishing clear governance will transform their service desk from a reactive support function into a proactive, strategic enabler for the entire organisation.