Meta Platforms has acquired the front page of the chatbot internet, Moltbook, a social network only accessible to bots.
The team behind Moltbook will join Meta’s Superintelligence Labs, which is responsible for training and operating Meta’s Llama AI model. It is unclear whether Moltbook will continue to function as an open platform or whether Meta will take it offline and use it for its own research purposes. The terms of the deal were not disclosed.
Moltbook took the internet by storm in late January, with users finding agents posting threads on religion, identity, and ethics. There has been some scepticism regarding the posts on Moltbook, as testers have said there appears to be an active human community pretending to be bots on the social network.
It is a big win for Matt Schlicht, the co-creator of Moltbook, who said he vibe-coded the entire website in a weekend. He and co-founder Ben Parr will be joining Meta in the next few days once the deal closes.
Meta has not been shy about using M&A to boost its AI ventures. It spent $2 billion on AI startup Manus, which added technical knowledge and AI agent services that are planned to be integrated into Facebook and WhatsApp.
These investments are seen as necessary by some commentators, who believe Meta is behind OpenAI, Anthropic, and Google on AI model sophistication and usage. It recently lost its head of AI, Yann LeCun, who has since launched an AI startup with a $1 billion seed funding round.
That said, Meta has been quick to integrate AI across its family of apps and is branching into new markets with the recent launch of a shopping AI tool.
Meta’s huge AI investment
The size of Meta’s investment in AI has troubled some, as the social networking company has a different structure to other giants such as Google and Microsoft, which have large data center operations, and is not a pure-play AI company like OpenAI or Anthropic.
In the company’s full-year financial call, it said it would double spending to $135 billion in 2026, almost all of which will go toward AI projects and related infrastructure. This is a huge increase, but deemed necessary by Meta as the total spend of Meta, Microsoft, Google, and Amazon this year is expected to reach $650 billion.
What’s not clear is the roadmap for how Meta will make this profitable. Improving ad sales revenue through AI is already seeing some progress, but with over $100 billion in spend, Meta would need Llama to be competitive in both the consumer and enterprise markets, neither of which is true at the moment.
Similar to Elon Musk’s xAI, the company is burning through money at an incredible rate, but does not have the grand vision of OpenAI or Anthropic, which are launching specific models and agents for different market sectors, alongside research and shopping improvements to make them the next version of search.
AI, smart glasses, but no more metaverse
Even though a significant portion of Meta’s spend is going on AI, the company is also expected to ramp up production of its smart glasses. Sales of its smart glasses tripled in 2025, driven by a wave of consumer interest in the devices for video recording.
Meta sees this as an AI play too, as users can access Meta AI through the glasses. In future versions, AI is expected to take a more central role in the experience.
What’s not getting any attention is the metaverse, which was the original pivot from CEO Mark Zuckerberg in 2021. That also received a ton of investment from Meta, but the usage and revenue from the company’s Meta Quest virtual reality headsets and metaverse software were not strong enough, forcing it to pivot again to smart glasses and AI.
Also read: Meta’s infrastructure push also includes a potential AMD chip pact worth up to $100 billion as it scales compute for its AI ambitions.
