On December 29th, 2025, Meta acquired Manus AI: a general artificial intelligence agent that can perform multi-step tasks ranging from research to analysis. Users find that Manus AI can create complete projects from start to finish without requiring intermittent prompts. The Singapore-based startup launched in March 2025 and achieved over $100 million in annualized average revenue in the eight months prior to the acquisition. While the exact transaction amount is not public, Manus AI was in a VC funding round at 2 billion dollars, so Meta must have paid at least that amount. 

This acquisition is one of Meta’s many initiatives to stack up against its peers in the AI space, all of which reveal a problem: the company is spending aggressively on artificial intelligence without a clearly defined strategy for how these technologies will generate revenue.

After all, Meta AI, Meta’s proprietary AI model, appears to lag behind other AI models. It receives far less attention as its OpenAI, Google, and Anthropic counterparts and lags in reasoning, software development, and other functionalities. While Meta AI had over 1 billion active monthly user base by mid-2025, that figure matches ChatGPT’s weekly user count and is less than 30% of Meta’s overall user base across its other apps. 

Importantly, Meta AI was never built to be ChatGPT. Today, it is an AI-powered virtual assistant for content creators. While this can be accessed through a standalone app, most users access Meta AI through Meta’s family of apps: Facebook, WhatsApp, Instagram, etc. Meta intends to integrate Manus AI features into Meta AI. Perhaps this integration will strengthen Meta AI’s capabilities and increase its versatility as Meta plans to begin testing paid versions of its currently free AI over the coming months.

Meanwhile, Manus AI will continue operating independently from Singapore, enabling users to continue licensing Manus AI separately from Meta AI. However, some of Manus’s existing users are upset about the shift, worried about Meta’s data usage policies and ability to innovate at Manus’s pace. 

While Meta is excited about absorbing Manus's talent, it has no shortage of talent in its AI division. After all, last year, it spent months and billions of dollars poaching from competitors and offering $100 million sign-on bonuses to leaders in the space, most notably spending $14 billion to secure a 49% stake in Scale AI and hire its CEO, Alexandr Wang. 

Meta is also actively investing in AI infrastructure. In mid-March 2026, Meta published a blog article about its progress and goals for Meta Training and Inference Accelerator (MTIA), a family of AI chips the firm is collaborating with Broadcom to produce. Meta has already announced the expected launch timeline for the next four generations of the MTIA over the next two years, committing to an iterative approach to avoid the lag caused by trying to catch up to technology after new breakthroughs emerge.

These investments in AI are driving significant increases in capital spending, which doubled from 2024 to 2025 and is projected to double again to $135 billion in 2026. Meanwhile, projected revenue for this year is only 50% more than its 2024 revenue. While Meta’s growth is huge for a company of its size, its revenue growth cannot keep up with its capital investments. The spending, therefore, is not to drive growth today but to drive future growth, signaling that Meta is committed to remaining at par with its peers moving forward. 

But what is Meta investing in? While users think of the firm as a social media company, it is in the advertising business. 97% of Meta’s 2025 revenue of $201 billion came from advertising. Many other technology firms are using AI to diversify revenue streams and if the acquisition of Manus is meant to be a standalone investment into a growing company, it may pay off once users are convinced that Meta is capable of protecting their data and growing the AI’s functionalities. However, beyond a standalone investment, it is difficult to see how Manus AI might fit into Meta's existing revenue structure. For comparison, Meta acquired Moltbook, a social networking platform for AI agents, in early March 2026, an acquisition that seems to align more closely with Meta’s mission.

Meta’s core challenge is strategic clarity and consistency. In a world of AI firms that have established goals and niches, Meta is trying, and failing, to do everything. It is copying and acquiring competitors, but until the firm clarifies its own goals and innovates to meet them, it will continue to lag behind leading AI models.