AI
direction, with 99% of IT leaders comfortable delegating at least one IT task to AI. Te most common include system updates (61%) and common device fixes (53%). Tis is only expected to increase, with IT leaders predicting that
nearly 40% of digital workplace services will run autonomously by 2030. As more tasks move from recommendation to execution, organisations will need to decide how much responsibility they are comfortable with giving to AI systems.
Trust will determine how quickly autonomy scales. People already recognise the value of AI, with 97% of workers reporting at least one workplace benefit. However, their confidence is tested when the technology moves from providing an output to taking action. More than half of employees (56%) say they oſten or always
verify AI-generated outputs before relying on them, spending an average of two hours each week doing so. Meanwhile, only 5% have no concerns about AI operating without a human in the loop. If employees are still checking AI outputs, handing it responsibility for making changes independently will require a greater degree of confidence. For the channel, this means looking beyond the level of
autonomous action an AI solution promises. When evaluating partners and solutions, channel leaders need to understand how vendors demonstrate trusted reliability of the solution, explainability of actions, and transparency behind decision-making. A solution may offer significant levels of automation, for example,
but partners will need evidence that it performs consistently in the environments where customers intend to use it, with clear visibility into why it took specific actions and what impact those actions had. Te ability to demonstrate and build trust in AI will become an important part of how partners assess AI vendors.
Guardrails need to become part of the AI conversation As AI systems gain access to more applications and data, the conversation around governance needs to evolve with them. Businesses need to determine which actions can happen automatically and where approval is required. Te appropriate level of oversight will vary according to risk.
Resolving a routine device performance issue has very different consequences from making a significant security or access change. A graduated approach gives businesses scope to expand safely.
Low-risk, repeatable tasks can be handled within predefined parameters, while more consequential decisions should continue to involve human judgement. Te level of oversight can then reflect the potential impact of each use case. Tese considerations will increasingly shape how channel leaders
evaluate which AI solutions they choose to take to customers. Being able to explain safeguards and implement policy guardrails will help partners have more informed conversations with customers about where they are comfortable giving AI greater responsibility.
The channel can turn trust into an implementation strategy Channel partners have an important role to play in helping end-user
www.pcr-online.biz
“A graduated approach gives businesses scope to expand safely.”
customers put these systems into practice. Teir role goes beyond simply implementing the technology, and increasingly, they can help customers determine where autonomy is appropriate and what evidence would justify expanding its role. Partners are well placed to make these judgements because they
understand their customers’ individual best practices, infrastructure and operational priorities. Tis gives them an opportunity to translate broad ambitions around AI into specific use cases where it can deliver measurable value without introducing unacceptable risk. For example, a channel partner could work with a customer to
introduce AI within a limited part of its IT environment and agree upfront how its performance will be assessed. Tat might involve tracking the accuracy of its decisions, the number of cases that still require human intervention or whether it delivers the expected improvement for the IT team. Partners can also bring the expertise and capacity to carry out this evaluation where customers may not have the necessary skills in-house or the bandwidth to manage it themselves. Tose results give the customer evidence on which to base the next decision about AI’s role. Te channel can therefore help customers treat autonomy as
something that develops through experience. As the technology demonstrates its value and reliability in practice, partners can use that evidence to advise on where its role can expand and where greater oversight is still appropriate. Crucially, the opportunity for channel leaders is to help customers
build autonomy at the pace of trust, rather than treating trust as something that has to exist before the journey can begin.
What this means for IT teams Human oversight will continue to play an important role as AI takes on more workplace tasks. Only 2% of workers believe AI might eliminate their role by 2030, while 33% expect it to assist them while they continue performing most tasks. Te more immediate change for IT teams will be in how their
time is spent. As AI takes on routine interventions, IT professionals can spend more time addressing complex problems and applying their expertise where context and strategic judgement are required. Te division between tasks that AI can handle independently
and those that still require human involvement will differ between organisations. As AI becomes capable of doing more, IT teams will ultimately need to determine where greater automation can genuinely improve the way they operate and where their own expertise remains essential.
September/October 2026 | 23
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52
