Prompt and Model
In practice

Meta launches Muse, a consumer-focused AI

Meta unveiled its Muse AI agent at Connect, positioning it as a Tamagotchi-style personal assistant for everyday users.

Meta unveiled its Muse AI agent at Connect, positioning it as a Tamagotchi-style personal assistant for everyday users

Meta introduced its new AI agent, Muse, at the company's annual Connect event. CEO Mark Zuckerberg presented the consumer-focused product, described as a Tamagotchi-style personal assistant designed for deep integration into daily life. Meta plans to push AI features everywhere across its products, insisting Muse is for adults only. The agent has drawn strong reviews from early adopters and analysts following its high-profile debut.

Early user tests show promise but reveal trust and functionality limits

Early testing reveals both potential and limitations. One user, Sean O'Kane, tried Muse and found it located unclaimed money for him, resulting in a check arriving by mail. He described the feature as a one-time, non-repeatable party trick. Upon download, Muse did not instantly connect to Threads, Instagram, or Facebook, initially working with him as if he were a stranger-a factor that increased his initial willingness to use it.

Over time, however, the agent attempts to pull in user data from Meta's platforms to learn about the user. O'Kane expressed distrust in Meta handling sensitive information, citing its ad-based business model. He compared Muse to Apple's Siri, stating he trusts Apple more with such data because its business is not selling ads. While he pondered using Muse for managing real financials like credit cards to cancel subscriptions, the trust concerns with Meta made that prospect unlikely. He found the agent lacked persistent utility beyond singular features.

Meta’s consumer AI bet contrasts with rivals’ enterprise focus

Meta's launch of Muse marks a distinct strategic path. While other major AI companies like OpenAI and Anthropic are focused on coding and enterprise tools, Meta is doubling down on consumer-friendly AI. Industry observers note a broader shift toward enterprise AI for monetization due to high costs. Meta may be pursuing a different opportunity by leveraging its embedded user base.

The company's strength lies in its everyday integration via Facebook, Instagram, and WhatsApp, products used by 43% of people worldwide. This massive, conditioned-to-share user base is seen as Meta's controversial asset and potentially its strongest argument for Muse's success. The agent itself functions as a chatbot interface on iPhone or Android, capable of suggesting actions like controlling a computer or managing finances. It represents Meta's ground-up version of an on-phone app.

Trust remains the central hurdle for widespread adoption

Despite the promising launch, trust is the central challenge. The key question is whether consumers will trust Meta with access to wallets, calendars, inboxes, and shopping habits. Zuckerberg has been repeatedly summoned to Washington over issues ranging from the Cambridge Analytica scandal to antitrust concerns and online child safety. He argues that privacy and security were built into Muse's systems from the ground up and that adoption depends entirely on user trust.

Zuckerberg has stated that the industry is entering a phase where trust and alignment are the most important next set of capabilities. Meta shares climbed almost 13% over the past week, and the company currently holds the market largely to itself in the AI agent category. Being first with a capable product may establish a dominant foothold before rivals like Google or an enhanced Apple Siri arrive. Meta's challenge is whether consumers will trust it with access to wallets, calendars, inboxes, and shopping habits.

Related coverage

More from In practice