Microsoft's Copilot struggles to make headway against rival AI assistants; Sensor Tower says Copilot's mobile app has 79M downloads, far below ChatGPT's 900M+
Tyson Jominy routinely summons Microsoft Corp.'s AI assistant by tapping the Copilot key on his computer. Not because he means to.
Context & Ripple Effects
Microsoft’s Copilot rollout has already shown a gap between distribution and sustained use: an internal presentation said weekly usage had been flat, while customers described Copilot deployments as requiring data preparation rather than working out of the box. Flat Copilot usage and deployment friction provide context for the mobile adoption shortfall.
The story matters because Copilot is being positioned across Microsoft’s products, yet related reporting found enterprise employees often preferred ChatGPT, making user preference—not merely access—the central competitive issue. Enterprise demand for ChatGPT adds to that pattern.
First-order effects
- Microsoft faces a clear mobile reach deficit versus ChatGPT, limiting Copilot’s ability to build consumer familiarity and repeat use through its standalone app.
- Accidental activation of the Copilot key can make prominent hardware placement feel intrusive rather than useful, putting more pressure on Microsoft to improve the interaction and perceived value.
Second-order effects
- The adoption gap reinforces the challenge for Microsoft sales teams: enterprise buyers may encounter employees with established preferences for rival assistants rather than a ready-made Copilot user base.
- Rival assistants benefit when consumer-scale usage becomes a workplace preference, raising the bar for Copilot to convert Microsoft’s product distribution into active engagement.
Third-order effects
- The result points to a broader test for the assistant operating layer: default placement and bundling may create awareness, but durable leadership depends on whether users choose the assistant repeatedly.
- If this pattern persists, enterprise AI competition will be shaped less by who embeds an assistant most widely and more by who earns cross-context user habits and trust.
The trend: AI assistants are shifting from a distribution race to a habitual-use race, where embedded access alone does not ensure adoption.