Meta Platforms (NASDAQ: META) took a real step forward with its artificial intelligence (AI) strategy in April. The company announced the release of its Muse Spark model, helping shares of the tech stock go on a solid run.
This was for good reason, as key AI evaluation platforms showed that Muse Spark is much more intelligent than Meta’s past Llama models.
On the other hand, the Magnificent Seven company has also taken a significant blow when it comes to another AI vertical.
Near the end of 2025, Meta announced that it would acquire the AI startup Manus. The deal made a lot of sense, with Manus’s capabilities synergizing well with Meta’s business offerings. However, months later, the Chinese government opted to block the transaction—putting a blemish on Meta’s agentic AI strategy.
The Manus Goal: Expanding Meta’s Automated Solutions
Manus, once a Chinese company, had quickly built up a large pool of users deploying its agentic AI solutions. This allowed the firm to scale its annual recurring revenue to $100 million just eight months after launching. Manus claimed that this was the fastest pace at which a company had achieved this feat. Manus’s success and the way its technology could improve Meta’s products enticed the company to purchase Manus for over $2 billion.
The presumed logic behind the Manus acquisition is relatively straightforward. Advertising drives nearly all of Meta’s revenue, with the company providing various tools to advertisers to support their goals. This includes Advantage+, which uses AI to automate and optimize advertising campaigns. The primary goal of this is to help advertisers get the most business in return for each dollar they spend on Meta. The better the company can do this, the more likely advertisers are to spend money on Meta.
Like Meta does for advertising campaigns, Manus helped automate other key tasks that businesses engage in, like market research, coding, and data analysis. By adding Manus's AI agents, Meta could offer a broader suite of tools to customers, automating advertising as well as these other functions. As the company already has a massive base of advertising customers, it is reasonable to think Meta could upsell them on additional automation products.
However, it appears that the Chinese government has dashed nearly all hope that Manus will become a part of Meta.
China Shuts Down Deal—Manus Moves to Comply
The Chinese government has blocked the Manus transaction, ordering Meta and Manus to unwind the deal. Although Manus moved its headquarters to Singapore, China claims that because the company had its beginnings in China, it still has jurisdiction over approving a sale of the firm. The decision signals that China will go out of its way to ensure that homegrown AI startups do not fall into the hands of American companies. In a strict legal sense, it’s not clear that Manus and Meta actually have to listen to this order. Additionally, many investors in Manus had already received payment before China blocked the deal, complicating matters in unwinding it.
However, Manus is nonetheless taking steps to move in that direction. Manus’s founders are aiming to raise $1 billion in order to buy back their stakes in the company from Meta, according to reports. Meta itself has not made any comment on the deal’s denial.
Even if China doesn’t have legal authority over Manus, there are likely forms of indirect power that it can use to get Meta and Manus to comply. For example, although China bans Meta’s apps, Chinese advertisers can still buy Meta ads in other countries. Thus, Meta’s core business still generates a lot of revenue from Chinese customers, making thwarting the government’s order a risky proposition.
It seems best to assume that Manus will not be part of Meta going forward. Still, Meta has yet to say the deal is off, suggesting that there may be some hope left when considering the complicating factors in unwinding it.
Meta’s Agentic AI Journey Likely to Progress Without Manus
Overall, the Manus deal having fallen through sets Meta back on its AI product release roadmap. This is particularly relevant, as Manus was one of the key examples of how Meta was pursuing agentic AI. Still, Meta has discussed agentic AI in many other contexts, including business messaging agents and personal agents. This is to say that Meta was not reliant on Manus to provide agentic solutions, although Manus could have helped accelerate the deployment of its agentic solutions.
Additionally, at a certain level, Advantage+ likely already uses some form of agentic AI, or a least functions closer to an agent than a chatbot. The automation of multi-step tasks—as Advantage+ does in setting up, running, and optimizing advertising campaigns—is one of the key things that distinguishes agents from chatbots. Thus, there is reason to believe that Meta can effectively develop other agent-like capabilities without Manus.
The article "Meta's Manus Mess: Why China Blocked the Deal and What It Means" first appeared on MarketBeat.
