Meta's long-awaited AI model is finally here. But can it make money?

Meta has officially launched its first AI model, Muse Spark, nearly ten months after securing Alexandr Wang from Scale AI to lead its AI initiatives. This move comes amidst growing concerns about whether users will be willing to pay for the new technology. While competitors like OpenAI, Anthropic, and Google have already established themselves with effective AI models and services, Meta's extensive investment in AI has yet to translate into new revenue opportunities. In June, the company invested over $14 billion in talent, establishing the Meta Superintelligence Labs as a premier unit focused on AI development. Moreover, Meta announced plans to significantly increase its capital expenditures this year, with projections between $115 billion and $135 billion, almost double its previous figures. Analysts have noted that the company needed to demonstrate tangible progress after a year marked by hiring rather than product launches. Morningstar's Malik Ahmed Khan emphasized that Meta must now focus on making the Muse Spark model functional and profitable. The launch of Muse Spark represents a shift from Meta's previous open-source Llama models. While the new model is proprietary, the company has hinted at the eventual release of some open-source versions. Zuckerberg's change in strategy follows a lackluster response to Llama 4, pushing Meta to pivot towards a more commercially viable AI offering. Looking ahead, Meta aims to provide third-party paid API access to Muse Spark following a private preview with select partners. However, the company faces stiff competition in a landscape dominated by well-established players like OpenAI and Google, who are already valued over $1 trillion due to their successful models and services. Analyst Andrew Boone points out that Meta's significant user base—over 3 billion across Facebook, Instagram, and WhatsApp—could be a key advantage. Rather than chasing developers who flock to competitors, Meta's strategy appears to focus on enhancing its core advertising business. Advertising accounted for 98% of its $200 billion revenue last year, and improving ad engagement through AI is seen as crucial for the company's future. Despite attempts to diversify into areas like the metaverse, Meta's ad model remains its most reliable revenue stream. As advertisers experience positive returns from their campaigns on Meta’s platforms, they may find value in AI services that further enhance their advertising effectiveness. Initial benchmarks for Muse Spark suggest it excels in image and video processing, making it particularly appealing to advertisers looking to create dynamic content tailored for platforms like Instagram Reels. However, some industry insiders express skepticism about the model’s ability to compete with both free alternatives and existing proprietary options. As Meta strives for AI independence, the company's massive investment in Wang's team raises expectations. The question remains: with a cutting-edge model now in hand, how will Meta capitalize on this opportunity?