Meta’s personal AI agent “Muse” gained millions of downloads within weeks of its September launch by promising to handle life’s annoying chores. This shifted the AI spotlight away from smartphone makers’ strategies. The rise of personal agents could reshape consumer device interactions, spending habits, and how much control device manufacturers retain over customer relationships, says GlobalData, a leading intelligence and productivity platform.
Apple, Google, and Samsung have already staked their AI claims, each using their flagship launches this summer to position generative AI as a new source of differentiation and loyalty by tying the technology to the operating system, apps, services, and connected hardware surrounding their phones.
Meta’s Muse arrives amid that summer push but takes a different route, running from a dedicated computer in Meta’s cloud and working from whichever phone a consumer happens to carry. OpenAI followed Muse with Dots, the company’s always-on agent, in late September and xAI has been testing Grok Bot since August.
GlobalData’s research reveals that personal AI agents are challenging the premise that the best consumer AI experience must be deeply integrated into the smartphone and connected across a wider device ecosystem.
The personal-agent shift extends to hardware as well, as companies begin designing devices around AI. Meta’s Muse Charm, a keychain-sized prototype targeted for December, strips away much of what consumers associate with a phone to focus on access to the agent, while Muse is also coming to Meta’s smart glasses.
Amazon, OpenAI, Nothing and Honor are exploring their own variations on AI-first hardware while Qualcomm is betting that more capable on-device agents will make the handset itself more important.
Amanda Mitchell, Senior Analyst – Telecoms Practice at GlobalData, comments: “What personal AI agents mean for hardware is far from settled. In one future, agents live mostly in the cloud and the phone matters less. In the other, agents need so much local context and processing that the device becomes more important than ever. The agent may be slipping free of the phone, only to come back looking for hardware of its own.”
Personal AI agents also open new monetization avenues that could shift more consumer spending away from smartphone makers’ ecosystems. Integrations with Walmart, Sephora, Expedia, OpenTable, and Shopify let Muse shop and book on users’ behalf, and Meta has said it expects a cut of transactions completed through the agent, giving the company a direct role in commerce without owning the smartphone or operating system the consumer is using. Muse also offers a free tier alongside paid plans at $20 and $100 a month, adding a subscription revenue stream geared toward heavier users.
Mitchell continues: “A risk for phone makers is that more of the consumer’s attention, decision-making, and spending could be captured and monetized by someone else’s AI, even if consumers keep buying smartphones.”
For consumers, the trade-off grows as AI moves from answering questions to acting on their behalf. Phones, chatbots and a growing wave of personal agents all want access to email, calendars, logins, payment details, contacts, and other personal context, raising obvious privacy concerns. At the same time, consumers may find themselves juggling multiple AIs, each with its own app, account, permissions, connections, and payment information. As those choices multiply, so do both the privacy decisions and the friction of managing them.
Mitchell concludes: “The promise of AI in your pocket is convenience, but every new assistant asks consumers to decide what to share, who to trust, and whether another layer of setup is worth it.”
