Muse AI, Meta's (META) new personal AI agent, has been downloaded about 1.1 million times since launch, still US-only, so that number should speed up. This pace is ahead of the Meta AI app but well behind Threads. When Threads launched in 2023, users could immediately link their Instagram accounts, which drove its early surge. Muse's downloads may also be held down because people can connect to it through WhatsApp instead of installing the app. Muse is doing better than Meta AI did at the same early point, but not as strong as ChatGPT (OpenAI) in its first week.
Does Muse change the Meta AI thesis? Maybe not yet. It is early days.
What Muse actually is
Muse is a personal agent, not a chatbot. It helps complete personal tasks - planning a trip, sending emails, updating a calendar - rather than just answering questions. The more personal data you connect, the more it can theoretically do for you. That makes Muse a closer match to Gemini Spark (which sits inside Gemini, so no separate numbers exist) or to a newer personal AI agent called Instinct, both of which handle daily-life tasks.
Why adoption is faster than expected
Two forces are at work. People are now much more comfortable with AI and putting it into daily life, and they understand better what it can do. Meta (META) also has a large existing customer base it can target over time. Privacy was a big consumer worry, so Muse runs as a separate private browser, and what you put into it is kept out of the Meta AI ad system. Beyond that, the broader AI adoption cycle carries it: everything is now AI. Claude (Anthropic) is being built into workspaces, and ChatGPT (OpenAI) is the largest app by monthly active users worldwide. People treat Muse as one more way to fold AI into everyday life.
Downloads versus the metric that matters
If the same features are available through another app like WhatsApp, that artificially lowers the reported download numbers. Downloads still matter for gauging broad customer adoption, but for a product with free and paid tiers, engagement is the real signal. A heavy user is likely to pay for the top subscription at $100 a month, versus $20 a month, versus free. So the numbers to watch are sessions and time spent, measured after the app has run for a while, and any rise or shift in them.
The core idea is stickiness. The more integrated an app is with a user's daily activities, the less likely they are to switch, even to another free or freemium product, because they have already connected their email, connected their calendar, and added a credit card. For a plain chatbot, a basic user often has no preference between Claude and ChatGPT - they just ask a question, and switching costs almost nothing. Once an agent is integrated into someone's life, the barrier to switching rises sharply. Engagement compared against other personal AI agents 6 months to a year from now will show whether Muse holds users.
