Analyst(s): Olivier Blanchard
Publication Date: May 9, 2025
Microsoft’s Copilot+ ecosystem initially looked like a unifying framework for AI PCs: It created order, predictability, and direction for AI PCs when the segment most needed it. Moreover, branding the category “Copilot+” effectively established Microsoft as the overseer of the nascent AI PC market. But now, less than 12 months after launch, the relevance of the Copilot+ category within the broader AI PC segment is starting to feel like it is losing steam.
Key Points:
- One critical reason why the Copilot+ PC category is at risk of losing its relevance in the still-nascent AI PC segment is the relative absence, still, of compelling on-device AI experiences for users: The vast majority of “visible” and agentic AI use cases continue to be cloud-based.
- Microsoft still struggles to scale the adoption of its Copilot assistant, which, despite its natural integration into the Windows ecosystem, is failing to move users from more established options such as ChatGPT and Google’s Gemini.
- Because Copilot+ class AI PCs are still neither defined by on-device AI experiences nor by Copilot as the de facto AI assistant, the relevance of Copilot+ as an AI PC category could begin to erode in the next year.
Overview:
A year ago, Microsoft launched its Copilot+ AI PC ecosystem in collaboration with Qualcomm and announced impending integration with AMD and Intel. It seemed like the right unifying framework for AI PCs. It established minimum performance standards, provided clarity of direction for the nascent market, and made the case for a more AI-forward Windows ecosystem that would help guide the PC market into the age of assistants and agentic AI.
PC Original Equipment Manufacturers (OEMs) aligned themselves with Microsoft’s vision, launching their first wave of Copilot+ PCs, which almost immediately established the category as the standard for Windows 11 AI PCs. But less than a year after its launch, the significance of Copilot+ as a cohesive ecosystem feels like it is losing traction. A primary reason is the persistent lack of compelling on-device AI experiences for users: while capable of handling decent AI workloads (both training and inference) directly on device without a connection to the cloud, the majority of visible AI applications, namely AI assistants and agentic AI, remain essentially cloud-based.
To make matters worse, Copilot – Microsoft’s AI assistant solution and the Copilot+ PC category’s namesake – has thus far failed to attract users at the same pace as competitors such as ChatGPT and Gemini. For reference, ChatGPT boasts 600 million monthly users. Google’s Gemini, which is integrated into Google’s productivity solutions such as Docs, Slides, Sheets, and Gmail, as well as Google’s Chrome browser and the Android mobile ecosystem, is already moving beyond 350 million monthly users. Meanwhile, Copilot, despite being integrated into Microsoft’s Windows ecosystem and having a dedicated prompting key on Copilot+ PC keyboards, is estimated to have fewer than 50 million monthly users.
Figure 1: Copilot Monthly Active Users Compared to Gemini and ChatGPT
Copilot’s inability to win over users quickly enough to compete against incumbent solutions could spell trouble for the relevance of the Copilot+ PC nomenclature in the long term, as well as for Microsoft’s AI assistant and agentic AI services play. The absence of a compelling reason for users to choose Copilot – or subscribe to Copilot Pro – as their primary AI assistant when alternative, potentially more familiar or preferred options already exist is a serious problem for Microsoft.
To add to the Copilot+ ecosystem’s challenges, the majority of AI PCs sold in recent quarters still fall outside of the category, making a strong case for the continuing mainstream relevance of systems shipping with sub-40 Trillions of Operations Per Second (TOPS) NPU: So long as AI assistants and agentic experiences continue to be mostly cloud-based, why spend more on a Copilot+ PC when a more volume-priced 20-39 TOPS system will do just fine?
If that were not worrisome enough for Microsoft, PC OEMs and their silicon vendors are also racing to put the category’s minimum system requirements in the rear view mirror. AMD notably pushed to 55 TOPS (NPU) within weeks of the category’s first releases, and upcoming processor refreshes look to be shooting for 100-200 total system TOPS for premium and flagship PCs by next year. Recent conversations with several vendors in the AI PC ecosystem touched on the possibility of >1K TOPS (total system) by 2030.
All of these details effectively chip away at the notion that the Copilot+ PC category will remain relevant in the long term, at least in its current form: If Copilot cannot be a key player in the AI assistant and agentic space in the Windows PC ecosystem, and, additionally, if hardware vendors in that same ecosystem continue to not only find their volume outside of Copilot+ but also leapfrog over minimum Copilot+ system requirements with hardware that is vastly superior to Microsoft’s framework for the category, does the category need to exist at all?
The full report is available via subscription to Futurum Intelligence’s Intelligent Edge and AI Devices IQ service—click here for inquiry and access.
Futurum clients can read more in the AI Devices Intelligence Platform, and non-clients can learn more here: AI Devices Intelligence Practice.
About the Futurum AI Devices Practice
The Futurum AI Devices Practice provides actionable, objective insights for market leaders and their teams so they can respond to emerging opportunities and innovate. Public access to our coverage can be seen here. Follow news and updates from the Futurum Practice on LinkedIn and X. Visit the Futurum Newsroom for more information and insights.
Author Information
Research Director Olivier Blanchard covers edge semiconductors and intelligent AI-capable devices for Futurum. In addition to having co-authored several books about digital transformation and AI with Futurum Group CEO Daniel Newman, Blanchard brings considerable experience demystifying new and emerging technologies, advising clients on how best to future-proof their organizations, and helping maximize the positive impacts of technology disruption while mitigating their potentially negative effects. Follow his extended analysis on X and LinkedIn.
