(Reuters) – Meta Platform (Meta.O) is considering partnerships with rival Google (Googl.o) or Openai to enhance artificial intelligence capabilities in its application, citing people familiar with the conversation, reported Friday.
Meta Superintelligence Labs, leader of Meta's new AI organization, has considered integrating Google's Gemini model and described the conversational, text-based answers submitted to Meta AI as the company's main chatbot.
The discussion includes leveraging Openai's model that powers Meta AI and other AI capabilities found in Meta's social media app.
Dealing with external model providers such as Google and Openai is likely a temporary step to strengthening META's AI products until their models progress, the report says. The lab's prioritization is to ensure that the next-generation model, the Llama 5, can compete with its rivals.
Within the meta, information reports that the company has already integrated external AI models into internal tools for staff. For example, employees can code through the company's internal coding assistant using the human model, cites three people who are familiar with the issue, according to the report.
“We take all the approaches to building the best AI products, including building world-leading models, partnering with companies and open sourcing technology,” a spokesman for Meta said in a statement.
Openai, Google and Microsoft (MSFT.O) support Openai and did not immediately respond to Reuters' request for comment. Reuters could not immediately review the report.
Earlier this year, Meta led billions of dollars to former scale AI CEO Alexandr Wang and former Github CEO Nat Friedman into the co-lead meta super intelligence lab, offering a generous compensation package to attract dozens of leading AI researchers to the initiative.
Related Topics
'; r_text[1] = ''; r_text[2] = ''; r_text[3] = ''; r_text[4] = ''; r_text[5] = ''; r_text[6] = '';var i = math.floor(r_text.length * math.random()); document.write(r_text[i]);
