Meta Could Drive AI Inflection Point After $2 Billion Move

We have another M&A transaction to discuss on Tuesday, and yes, it is AI-related, as Meta (META) has agreed to acquire Singapore-based AI startup Manus in a deal estimated to be more than $2 billion.

Granted, that is a relatively small amount given the cash on Meta’s balance sheet, but it would bolster Meta’s effort in the area of AI agents as it incorporates Manus’ general-purpose AI agent across the company’s consumer and business products.

AI Agent vs AI Model

Before we go any further, let’s take a step back and review the difference between an AI agent and an AI model:

AI models primarily generate text or images (generative AI)
An AI agent is a software program capable of acting autonomously to understand, plan and execute tasks

Essentially, it boils down to knowledge versus action, with some comparing an AI model to the “brain” that processes information, while an AI agent is the “worker” that uses the brain to autonomously perform tasks.

We’ve discussed AI adoption and usage quite a bit in recent months, and while we’ve heard companies like Salesforce (CRM) , Perplexity and others talk about AI agents. Deloitte's 2025 Emerging Technology Trends study, which surveys 500 U.S. tech leaders, found that 30% of the surveyed organizations are exploring agentic options, with 38% piloting solutions and only 14% having solutions ready to deploy.

We are seeing the use of AI agents in several areas, notably healthcare, life sciences, finance, insurance, software development, IT and supply chain operations.

Could 2026 Be an Inflection Point for AI Agents?

Possibly. Gartner predicts that 40% of enterprise applications will feature task-specific AI agents in 2026, compared to around 5% in 2025. As we’ve shared before, we tend not to focus on the actual forecast figure, but more on the vector and velocity it reflects.

This helps explain the motivation behind Meta acquiring Manus as it looks to shore up its competitive offerings to better compete in the evolving landscape. Manus also fits the longer-term goal outlined by Meta CEO Mark Zuckerberg: AI systems that can assist users in achieving complex objectives, rather than merely responding to prompts.

Our initial take is that this is a positive move by Meta as it looks to compete with the likes of Google (GOOGL) , OpenAI and others primarily in the enterprise space. What is also interesting is that Manus’ revenue stream is primarily a subscription one, which is quite different from Meta’s, which is 99% advertising-based.

On the surface, the Meta-Manus acquisition, much like Nvidia’s (NVDA) pick up of Grog, and ServiceNow (NOW) adding cybersecurity Armis, is a step in the right direction. How big of a step will hinge on how well Meta can integrate it and use its AI agent to compete. That should become clearer as we move through 2026, but it gives us a reason not to count Meta out.

On Monday, we talked about CEO expectations for hiring in 2026. We continue to think the topic of AI adoption and usage will be a key theme when companies report their December quarter results and issue their initial take on what’s ahead in 2026. Helping set the table for that, AI and related use cases are likely to be a hot topic at CES 2026 next, especially with Nvidia’s CEO Jensen Huang and AMD’s (AMD) Lisa Su both giving keynote addresses on January 5.

Connecting the Dots, Part I

When we discussed the Nvidia-Grog and Softbank (SFTBY)-DigitalBridge (DBRB) deals on Monday, we said we would not be surprised to see the pace of M&A deals continue as companies look to plug geographic, product, or technology holes or improve their competitive positioning.

While we didn’t expect to see Meta join the list so quickly, it confirms our view that companies are jockeying to position themselves for what’s to come. We see that keeping an active M&A market, one that should benefit our shares of Morgan Stanley (MS) and Bank of America (BAC) .

Connecting the Dots: Part II

The adoption and usage of AI agents, which consume more data and resources than static AI models, and the same for AI-generated video, have the potential to congest digital networks and infrastructure.

We will want to closely monitor related figures and data points, including those from ServiceNow, Palantir (PLTR) , Microsoft (MSFT) , Google, Meta, OpenAI and others. If we see a dramatic increase in AI agent adoption, AI-generated video, or both, it would be a signal that we are likely to see capital spending increases to address related capacity utilization issues.

Beneficiaries of that spending would include Nvidia, Broadcom (AVGO) , Marvell (MRVL) and Arista Networks (ANET) .

More Pro Portfolio

At the time of publication, TheStreet Pro Portfolio was long META, GOOGL, NVDA, NOW, MS, BAC, PLTR, MSFT, AVGO and ANET.