Key takeaways: Meta's Muse agent app has reached the top of app store download charts, but it remains unclear whether users will integrate such digital assistants into their daily lives.

Like Meta, Marissa Mayer's Dazzle has also adopted a free strategy, observing user behavior while exploring viable business models.

IDC Chief Research Officer Meredith Whalen said in an email: "For Meta, the real core issue is trust."

On September 23, 2026, at the Meta Connect conference in Menlo Park, California, Meta Platforms CEO Mark Zuckerberg unveiled the Muse Charm hardware device.

Meta's personal agent Muse has topped app store charts across major platforms. The product, which Zuckerberg has called the "core pillar" of the company's AI strategy, now faces a key challenge: getting users to develop a sustained usage habit.

Market enthusiasm for Meta's AI agent has been high. Since its launch one month ago, Meta's stock has risen approximately 20% cumulatively, posting its best monthly performance since 2022 in September.

Although Muse has maintained high download numbers on Apple's iOS app store, the Facebook parent company faces a bigger challenge: retaining users after the novelty fades. Meta hopes this personal agent will break through in the fiercely competitive AI race — the company currently lags behind rivals such as OpenAI and Google in the AI space.

IDC Chief Research Officer Meredith Whalen said in an email: "Muse has an opportunity to capture users because it is a task-execution agent, not just a simple chatbot."

Whalen noted that users are already using various AI tools for diverse tasks, but it is uncertain whether they are willing to hand over the large amounts of personal data needed to build an effective digital assistant.

This obstacle is particularly prominent for Meta: the company has just reached massive settlements with multiple states that had accused Meta of deliberately downplaying the harm platforms like Instagram cause to teen mental health. "The real issue is trust," Whalen said.

Meta is promoting Muse to both ordinary consumers and small businesses, with capabilities including organizing personal emails, finding shopping deals, and identifying potential customers. Users can access Muse for free, while the platform also offers monthly subscription plans priced at $20 and $100 tiers based on compute usage — a metric Meta uses to measure the computing power required for its AI agent to complete tasks.

Whalen analyzed that offering free access in the early stage helps companies understand users' usage scenarios and motivations, and that "pricing discussions can come after users form habits."

Programmers have already found AI agents highly useful, but the general public and even business managers are still exploring the broader possibilities of powerful digital assistants beyond simple question-and-answer tools.

Ketan Karkhani, CEO of ThoughtSpot, which sells AI agent technology to enterprises, said: "Overall, the adoption of enterprise-grade AI agents has only just turned the first page of chapter one."

Consumer use of AI agents is also still in its infancy. Karkhani believes there is not yet a "killer application scenario" that would make products like Muse more appealing than next-generation Alexa or Siri voice assistants — which users often rely on for checking the weather.

Andrew Geller, executive producer at creative production company 1stAveMachine, said his company has developed internal agent tools to handle some production work. In his view, consumer-facing AI agents like Muse are "quite helpful" in popularizing the technology among the general public, but not life-changing.

Geller was an early user of an open-source AI agent orchestration framework that gained popularity earlier this year. He said that personal use of AI agents is indeed interesting, but "somewhat underwhelming" compared to high-intensity professional scenarios like coding.

Personal AI Assistant Race Draws Many Competitors

Multiple tech companies are commercializing AI agents for the general public, generally using free models first to attract users.

Former Yahoo CEO Marissa Mayer founded Dazzle, which focuses on photo AI agents and uses a try-before-you-pay model. She said this strategy is crucial for occasional users and first-time product triers.

Dazzle can analyze users' photos, identify personal information such as family members and hobbies, and then execute tasks — for example, reminding users that shoes they previously browsed are now on sale online.

This approach stems from her early years at Google, when the company had not yet built its now market-leading online advertising platform.

Speaking about Dazzle's first week, Mayer said: "We really want to figure out which scenarios resonate with users and which features they find genuinely useful, to guide subsequent iterations."

Mayer compared this pricing and user growth strategy to Uber's development: the ride-hailing company started with premium sedan services before gradually covering a broader mass market. "Being able to get quality recommendations and having this level of convenience is itself a kind of luxury," Mayer said.

Dazzle is considering two monetization methods: transaction commissions and subscriptions. "We want to build a service that makes users feel value continuously over the long term," Mayer said.

Wajo AI CEO Shivani Poddar is also exploring business models for her AI digital assistant startup, which supports users in researching information, booking services, and buying gifts for family members. Poddar previously developed voice assistant technology for Amazon Alexa and Facebook, and her product is currently free.

She believes enterprise customers would be willing to pay subscription fees, but is uncertain whether ordinary consumers would; regular people might be more accepting of result-based service fees, similar to Airbnb's commission model on bookings.

For Meta, the stakes are highest. Last year Meta recruited talent on a massive scale, aiming to build an AI business capable of competing with OpenAI and Anthropic.

Zuckerberg is eager to open up revenue sources beyond advertising — which still contributes nearly all of Meta's revenue; he needs substantial time to prove Muse can achieve profitability.

At least in the short term, Zuckerberg still has a window of time. IDC's Whalen said: "The party with the ability to subsidize users and wait patiently will have the advantage. Meta's advertising business is sufficient to support long-term investment and a slow wait for the product to mature."