Zuckerberg’s $2 billion AI shopping spree adds Manus, a Chinese-founded AI agent startup with buzz

Meta just made its boldest agentic AI bet yet

Meta is on an AI buying spree, and the latest pick is Manus, a startup with Chinese roots now based in Singapore.

Reporting by multiple outlets estimates the deal at roughly $2–3 billion, modest for megatech deals but large for such a young startup. To me, the message is simple. Meta wants AI that does things, not just AI that talks.

Manus is built to act like a digital coworker

Most chatbots wait for prompts and then respond. Manus is pitched as something closer to an autonomous agent that can plan, decide, and execute multi-step tasks with lighter supervision.

Think resume screening, trip planning, research workflows, and business back-office chores. If you have ever wished your assistant could finish the job instead of handing you homework, that is the promise Manus is selling.

Meta wants Manus inside the apps you already live in

The strategic logic is distribution. Meta can drop Manus-style agents into Facebook, Instagram, and especially WhatsApp, where small businesses already run customer chats all day.

Some analysts say Meta appears to be pursuing a WeChat style super app experience by expanding messaging into a place where users can accomplish practical tasks. Keep users in the app longer, then monetize that time through ads, commerce, and paid tools.

The buzz arrived fast, and the revenue claims are loud

Manus went viral online as a general-purpose agent, and the company says it is already generating real money through subscriptions.

Manus itself said it had crossed $100 million in annual recurring revenue and described a roughly $125 million revenue run rate.

Manus has a Windows desktop build listed in the Microsoft store and reporting indicates the product has been trialled on PCs in at least some contexts.

The startup’s origin story adds a geopolitical wrinkle

Manus traces back to the Butterfly Effect, also known as Monica.im, which originated in Beijing before relocating its headquarters to Singapore as it pursued global expansion. In today’s climate, that origin story is not a footnote.

It is the headline risk. Chinese founding ties can trigger suspicion about data access, influence, and control. Meta is buying capability, but also purchasing scrutiny.

Meta is promising a clean break from China-linked ownership

Meta said the company will discontinue Manus services and operations in China and that there would be no continuing Chinese ownership interests after the transaction, a statement intended to address regulatory concerns.

Observers argue the plan is intended to remove potential leverage points that might trigger foreign influence concerns among regulators and policymakers. Whether the structure satisfies Washington will depend on governance details and how access to systems is enforced.

Washington will review this deal in the same manner as it reviewed TikTok

Recent years have trained U.S. policymakers to treat China-linked consumer tech as a national security concern. Manus is a new test case in the AI era, and critics have already begun to circle it.

After a U.S. venture firm led a funding round, Senator John Cornyn blasted the idea of backing a potential adversary in AI. Expect tough questions about data, code, and control.

Meta’s bigger plan is to turn infrastructure into products

Zuckerberg has been clear that AI is the priority, and Meta has invested heavily in chips, data centers, and its Llama model lineup. The problem for any company is converting that spend into things people will actually use and pay for.

Acquiring Manus is a shortcut. Instead of building an agent ecosystem from scratch, Meta gets a product, a team, and momentum it can plug into its platforms.

This purchase matches Meta’s recent talent and tool-grabbing

Meta has been acquiring AI capabilities through deals and hires, including a significant investment in Scale AI that brought founder Alexandr Wang into a central leadership role. The Manus move is another attempt to accelerate growth through acquisition.

It is also a way to keep pace as OpenAI and Google race to own the consumer assistant layer. In AI, speed is strategy.

The integration phase is where shiny acquisitions get tested

Buying an AI startup is easy compared with making it feel native inside giant apps. Meta needs Manus to be fast, reliable, and safe across a massive scale.

It also has to avoid hallucinated actions, privacy mistakes, and creepy automation that users reject. Meta says it will keep the Manus service running while integrating roughly one hundred employees into its broader AI organization.

Agentic AI forces new rules about trust and permissions

When an agent can do more, it can also do more harm. That means tighter safeguards around permissions, data boundaries, and audit trails. Users will want clarity on what the agent can access, what it stores, and how it explains its actions.

For businesses, the bar is even higher, as a single mistaken automated decision can quickly result in significant financial losses. Functional autonomy needs visible guardrails.

The edge is shifting away from innovation alone toward execution

The next fight is not only about who has the biggest model. It is the one who has the most helpful agent wrapped around it. Manus claims it requires less prompting and can complete complex tasks, which is precisely what rivals are also pursuing.

Meta’s advantage is reach. If it packages agentic help inside WhatsApp and Instagram without friction, it can win on convenience, not benchmarks.

If you’d like a peek inside how those agent bets are playing out behind the scenes, please take a look at why big-money hires at Meta are stirring tension among AI researchers.

Users may get more intelligent helpers and bigger privacy debates at once

In the near term, you will likely see Meta assistants that can draft, plan, summarize, and execute across your chats and feeds. That could be great, especially for creators and small businesses.

However, you should also expect more prompts requesting permissions and more debate about what data is used to train and personalize agents. The acquisition is a bet that people want AI that acts on their behalf.

To understand how this shift fits into Meta’s broader strategy, it’s worth taking a quick look at why Zuckerberg is pivoting the company toward paid AI as its open-source approach wanes.

What do you think about Mark Zuckerberg investing in a major Chinese-founded AI startup now based in Singapore amid the AI boom? Please share your thoughts and drop a comment.

This slideshow was made with AI assistance and human editing.

Don’t forget to follow us for more exclusive content on MSN.

Read More From This Brand:

• Mark Zuckerberg signals the end of his metaverse ambitions
• Zuckerberg trims metaverse budget, eyes AI and hardware
• Mark Zuckerberg reveals Meta will increase AI content in your feed