Meta’s Muse App Gains Traction: What This Signals for Its AI Strategy

Meta Platforms (META) launched Muse, a personal artificial-intelligence agent, on Sept. 8. Within two weeks it had climbed to the top of Apple's U.S. App Store, and Wall Street analysts began treating it as the first clear evidence that Meta can turn its AI work into a product consumers seek out. The stock rose nearly 30% from the start of September through Sept. 22, MarketWatch reported, and on Oct. 7 its chart formed a golden cross, a technical pattern that marks an upward trend.

What Muse does and how fast it spread

Muse is built to carry out multistep tasks rather than only answer questions. It can book reservations, make phone calls, send emails and fill out web forms for its user. That places it among AI agents, software that acts on a user's behalf instead of only holding a conversation.

The download data shows how quickly it caught on. Muse reached 1.43 million cumulative U.S. downloads in the 12 days after launch, according to Sensor Tower figures cited by Bank of America analyst Justin Post. Over a comparable period after their own launches, OpenAI's ChatGPT had 1.37 million and Google's Gemini had 0.71 million.

JPMorgan analyst Doug Anmuth, who upgraded the stock to overweight in September after the launch, pointed in a Sept. 22 note to three advantages: a user-friendly interface, a generous free tier, and Meta's ability to distribute the app across its social-media platforms. "While it is still early, we believe that Muse has the potential to become the most widely used consumer AI application since ChatGPT," he wrote.

Why it matters for Meta's AI strategy

Until recently, investors had largely dismissed Meta's AI strategy. Post wrote that the market had doubted Meta's ability to build a differentiated AI product, and that Muse's popularity and its ability to complete complex tasks have spurred "optimism on Meta's potential AI investment returns."

That question matters because Meta's revenue still comes overwhelmingly from advertising across Facebook, Instagram and WhatsApp, while the company has been spending heavily to build AI models and the data centers that run them. A consumer product with this kind of early adoption gives investors a way to judge that spending other than by its cost.

Anmuth also noted that Meta has sped up the pace of its AI releases, and that its models have made large leaps in capability within a few months. The release of Muse Spark 1.1 in July was the first time Meta put out a model with near-frontier coding capabilities, he said.

The monetization gap

What Muse has not yet produced is revenue. Post cautioned that the app's popularity may not lift sales right away. "The monetization path for Muse, including ads, subscriptions and commissions, seems unlikely to be material before 2028," he wrote. On Oct. 7, MarketWatch reported that investors still had not identified how Meta will make money from the product.

That leaves the rally resting on expectations rather than reported results. Jefferies analyst Jeffrey Favuzza warned on Sept. 22 that enthusiasm ahead of Meta Connect, the company's developer conference scheduled for Sept. 23 and 24, could lead to a "sell-the-news event," after the shares jumped 11% on Sept. 21. The stock slipped 0.6% on Sept. 22 in up-and-down trading. Favuzza added that, with a new model release still ahead, "we have a more durable floor in the stock." Anmuth expected the conference to bring updates on how Muse would work with partners and with Meta's hardware.

What the chart and the calendar show next

A golden cross occurs when a stock's 50-day moving average rises above its 200-day moving average. According to Bluekurtic Market Insights, the Oct. 7 cross was the seventh in Meta's history, and after each of the previous six the stock was higher a month later. Both averages are built from past prices, so the pattern confirms a trend that is already under way.

The next catalyst investors are focused on is Meta's next AI model, code-named Watermelon. Analysts cited by MarketWatch on Sept. 22 expected it to be released in late October. Post said Watermelon could bring frontier-level capabilities, with the potential to improve Meta's core advertising business and to add revenue from hosting models.

What to watch:

• The release date and capabilities of Watermelon.
• Whether Muse keeps its App Store position and download pace after the launch period.
• Any figures on Muse usage, or on ads, subscriptions or commissions inside the app, in Meta's next quarterly report.