According to ZDNet, Meta announced on Monday it is acquiring the Singapore-based AI startup Manus for a deal valued at over $2 billion. The startup, which was originally founded in China, went viral back in March for launching what it promoted as the first truly useful general AI agent. Meta stated that Manus’s talent will join to build “general-purpose agents” across its consumer and business products, including Meta AI. The company, currently ranked as the sixth most valuable in the world, has been playing catch-up in consumer-facing AI despite its vast resources. This acquisition follows Meta’s reported $14.3 billion investment in Scale AI in June and the appointment of that company’s CEO as its chief AI officer.
Meta buys a shortcut
Here’s the thing: Meta has the users—billions of them across Facebook, Instagram, and WhatsApp. But it’s been lagging behind in making AI feel genuinely useful and integrated into daily digital life. OpenAI has ChatGPT, Google has Gemini woven into Search, and Anthropic has its claude.ai. Meta AI? For many, it’s still that thing you might accidentally trigger in a group chat.
So this move is classic Zuckerberg. Can’t build the leading tech fast enough? Buy the team that did. Manus made a splash because its agent could actually do things—browse the web, analyze data, fill out forms—with less hand-holding. That’s the holy grail right now. Integrating that capability directly into Instagram DMs or WhatsApp could suddenly make Meta’s apps feel a lot smarter. But it’s a huge integration challenge. Will it feel seamless, or just bolted on?
The two-track strategy
What’s interesting is that Meta isn’t just absorbing Manus. The startup will keep running and selling its own service. That’s clever. It means Meta gets the tech and talent for its own walled garden, and it gets a separate revenue stream from Manus’s existing and future enterprise customers. That cash can then fuel Meta’s even bigger, crazier ambition: being the first to build “superintelligence.”
Talk about hedging your bets. On one track, you’re trying to make a better chatbot for your social apps. On the other, you’re funding a moonshot. It shows how scattered Meta’s AI focus has been, pivoting from the metaverse to an “Efficiency” drive to now an all-out agent push. But maybe having Manus operate independently is the key. It keeps the startup’s culture and momentum alive, instead of drowning it in Meta’s bureaucracy.
What this means for everyone else
For users, the promise is that your apps might soon have a truly capable assistant built-in. Need to plan a trip with friends in a Messenger group? The AI could research flights, summarize options, and even book them. That’s the dream, anyway. The reality will probably be clunkier at first, and privacy questions will be massive. Do you really want Meta’s AI handling your sensitive tasks?
For the AI industry, it’s another sign that the agent wars are heating up. OpenAI has Operator, Google is building its own stuff, and now Meta has a $2 billion head start. It also highlights the insane valuation and competition for top AI talent. A startup goes viral in March and gets acquired for over $2 billion by September? That’s the market we’re in.
Basically, Meta is trying to buy its way to the front of the line. They have the scale, the data, and now a leading agent team. But execution is everything. Remember the metaverse? They had all the resources for that, too. Turning a viral demo into a reliable, billion-user product is a whole different game. I think the next year will show us if this was a genius acquisition or just another very expensive pivot.
