Why Meta Is Turning to Its Rivals to Win the AI Race

Meta is diversifying its AI approach, pursuing an “all-of-the-above strategy” that combines its own Llama models with external collaborations while evaluating temporary use of rival technologies. According to The Information, Meta’s newly established Superintelligence Lab is assessing whether to integrate Google Gemini or OpenAI’s models into its Meta AI chatbot and other services to boost conversational quality and text-based search.

Meta engineers already have access to Anthropic’s models to aid in software development, and the company is offering highly competitive compensation to attract AI researchers from Google and OpenAI, expanding the scale and expertise of the lab. In essence, Meta’s strategy favours flexible integration and parallel advancement, aiming to strengthen its ecosystem in the short term while building data and use cases to improve its own models for long-term competitiveness against Google, OpenAI and Anthropic.