Meta rolls out internal AI tool as it pushes into business market
Meta is advancing its efforts to develop artificial intelligence tools for businesses, unveiling an internal AI software designed to enhance productivity.
Called Metamate, named after Meta employees’ internal nickname, the software is built on the company’s large language model, Llama. It’s currently used for tasks like coding, conducting research, and drafting internal and external communications.
Prashant Ratanchandani, Meta’s Vice President of Engineering overseeing Metamate, stated there are no plans to release the software externally. However, the initiative provides valuable insights into enterprise AI use cases. He expressed Meta’s goal to create “the world’s best enterprise assistant.”
This move aligns with the broader race among tech companies like OpenAI and Google to introduce AI-powered tools for businesses and consumers, aiming to capitalize on the rapidly evolving technology.
The market for AI assistants capable of reasoning and performing complex tasks is forecasted to expand from $5.1 billion today to $47 billion by 2030, according to market research firm MarketsandMarkets.
Meta is also exploring monetization opportunities for its AI efforts through enterprise solutions. Recently, the company appointed Clara Shih, formerly of Salesforce, to lead a new “Business AI” division. Shih aims to make “cutting-edge AI accessible to every business.”
Analysts view Meta’s approach as a chance to commercialize Llama, potentially through ready-to-use applications leveraging its open-source models.
CEO Mark Zuckerberg has emphasized his commitment to AI dominance, integrating AI features into Instagram, WhatsApp, and Facebook while offering AI-powered tools for users, creators, and businesses.
Still, Metamate is less advanced than rival offerings from Microsoft and Google, which have internally-used tools already available to businesses. Unlike consumer-focused AI chatbots, Metamate is text-only and does not yet support video or image interactions. It also lacks autonomous features like scheduling or summarizing meetings, which competitors provide.
Meta employees, including executives, frequently use Metamate for various tasks such as coding, retrieving internal data, and preparing for client interactions. The company also uses other models, including OpenAI’s ChatGPT, for internal purposes.
Ratanchandani noted plans to make Metamate more assistant-like but said it is too early to evaluate its productivity impact.
Adoption of AI tools across the market remains limited. A Goldman Sachs report suggests widespread adoption and productivity benefits may not materialize until 2027.
Oren Etzioni, a computer science professor at the University of Washington, commented on Meta’s approach of testing its tools internally. While effective internally, he questioned Meta’s ability to compete as an enterprise vendor, given its advertising-centric focus as a social media company.