While the channel is increasingly optimistic about AI, that enthusiasm doesn't extend to AI PCs, according to new research from Canalys.
AI-capable PCs - desktops and notebooks that include a chipset or block for dedicated AI workloads such as a neural processing unit (NPU) - have seen a growth rate of 49% in the last quarter, the consultancy found.
They now make up 20% of PC sales, with Windows devices accounting for a 53% share.
More than four-in-ten channel partners told Canalys they identify AI as either a 'tremendous' or 'significant' growth opportunity.
However, more than around a third said they don't plan to sell Copilot+ PCs next year, with a further 34% telling Canalys that they only expect such devices to account for less than 10% of their broader device sales across 2025.
A quarter said announcements around AI-capable PCs had affected their customers’ refresh plans, with 11% noting it has accelerated plans. Yet despite this, 15% revealed they were being delayed until devices are available.
Notably, Canalys’ research found the underlying value propositions of AI-capable PCs aren't sufficiently clear to channel partners or their customers, partly because of their lack of availability.
"Despite the positive momentum, significant work must still be done to convince both channel partners and end customers of the benefits of AI-capable PCs," said Ishan Dutt, principal analyst at Canalys.
"This is especially true for more premium offerings such as Copilot+ PCs, which Microsoft requires to have at least 40 NPU TOPS alongside other hardware specifications.
"With the Windows 10 end-of-support now less than a year away, the coming quarters represent a critical opportunity to drive a significant portion of an aged installed base to be upgraded to an AI-capable PC."
PC vendors need to do more to equip, educate, and encourage channel partners to evangelize AI-capable PCs as part of their AI solutions, the consultancy said.
This means investing in education to teach partners about the benefits of AI-capable PCs, both for their own businesses and their customers, while giving them market development fund support and detailed knowledge about product roadmaps to help them better serve customers.
Training should involve first-hand experiences of effectively deploying on-device AI in their own organizations, Canalys said.
Meanwhile, there's a high level of user resistance to AI, largely thanks to security and ethical concerns, as well as a lack of trust between users and AI solution providers.
Overcoming this will require firms to incorporate privacy considerations at every stage of AI product development, from the training of models to how end-user data is accessed.
Alongside this, the consultancy recommended keeping communication clear and transparent and working with other industry players, governments and regulators to develop frameworks for the use of AI in consumer, business and education settings.
