Meta puts Muse AI at core of its strategy

Meta is placing its Muse AI agent at the heart of its long-term strategy, outlining plans to organise the company’s future around this new platform. The agent is positioned as the main product expression of an ambition to deliver personal superintelligence and is expected to serve a very large global user base over time. Muse has already achieved clear product market fit, and Meta is now shifting into a more operational phase focused on scaling the service. Executives have tied this vision to a broader AI expansion plan, including new funding structures and increased infrastructure spending.

Meta is introducing new non-voting shares to help finance its costly AI buildout, signalling that Muse and related capabilities will require significant ongoing investment. The company reported US$64.6 billion in quarterly revenue, up 26%, highlighting that its advertising business is currently financing much of this AI push. At the same time, Meta spent US$34.9 billion on capital expenditures in the quarter and generated negative free cash flow of US$2.1 billion, underscoring the financial scale of its AI ambitions. Management nonetheless frames Muse as central to Meta’s next phase of growth.

Muse is currently available in the United States and Canada, where more than 10 million people have set up accounts in roughly a month and a half. Meta has described adoption as strong, with early engagement more than twice the level reached by Threads at launch. Roughly 25% of Muse users are using the agent daily, indicating frequent interaction among a significant portion of the base. These early metrics have encouraged Meta to focus on scaling access, even as it manages costs.

Despite the early traction, Meta has delayed a global rollout of Muse, citing the expense of operating the agent at larger scale. The company has not disclosed per-user operating costs but has described compute demands as high. Management has said that more than a tenfold improvement in compute efficiency over the next year is needed to make Muse’s economics work at global scale. Until those improvements are realised, expansion will be paced against infrastructure and cost considerations.

Meta is extending Muse beyond its standalone app into dedicated hardware and wearables. The Muse Charm is a small device that offers real-time voice and avatar interaction without requiring a phone, giving users a portable way to access the agent. Meta is also preparing to launch Muse on its smart glasses, after previewing the integration during a recent event. On the glasses, users will be able to talk to their agent in a way similar to current Meta AI interactions, including checking on tasks or adding items to a to-do list.

Meta expects Muse to increasingly power its hardware strategy, shifting Reality Labs toward AI-powered wearables and away from a prior emphasis on virtual reality alone. Muse is also being integrated into Meta’s internal operations. The company is moving content-understanding systems onto its own models, which has reduced per-unit processing costs by more than 80% in offline evaluations. These moves show Muse functioning both as a consumer product and as a core technology underpinning Meta’s platforms.

Meta plans to monetize Muse through several channels, including collecting merchant transaction fees, selling subscriptions and offering enterprise AI services. The company has indicated that some Muse features, such as those on smart glasses, are likely to sit behind a paywall. Management has also said it expects to move to monetize Muse somewhat sooner than its usual product playbook, reflecting the agent’s higher operating costs. These plans are intended to support the heavy spending required for AI infrastructure.

Access to commerce platforms presents hurdles for Muse-related revenue. Amazon (AMZN) blocked Muse in September, citing security concerns, limiting potential transaction activity through that channel. Major financial institutions including JPMorgan Chase (JPM), Bank of America (BAC) and Charles Schwab (SCHW) are restricting outside agents from moving customer money, which could also constrain commerce flows. These conditions may limit the extent to which shopping activity around Muse can translate into transaction-based revenue.

Meta’s AI expansion and Muse rollout are taking place against a backdrop of significant legal challenges. A judge in New Mexico ordered a US$25 billion penalty after a jury found that Meta misled residents about how it shared user data with third parties and how it enforced policies on hate speech and misinformation. The case relates to activity involving British political consulting firm Cambridge Analytica. Commentators have described the penalty as a stark reminder of the legal pressures facing the company, which are expected to continue into next year.

These legal issues add another layer of risk as Meta reorganizes around its AI agent business. While Muse is becoming central to the company’s product lineup and internal systems, Meta must manage substantial capital spending, high operating costs and regulatory scrutiny at the same time. The combination of rapid AI deployment, ambitious monetization plans and large financial and legal obligations defines the context for Muse’s next phase of growth.