According to multiple sources, Meta is accelerating its expansion into artificial intelligence by establishing a new research team named "MRS Research," dedicated to optimizing its core recommendation algorithm system. The team falls under Meta's Recommendation System (MRS) department and drives content distribution across feeds in apps such as Facebook and Instagram. Records indicate that MRS Research was formed during an internal restructuring in October 2025, aiming to consolidate relevant research capabilities and advance long-term AI breakthroughs. The group will focus on cutting-edge research and academic publications, targeting a "leapfrog enhancement" of current recommendation systems while collaborating closely with advertising divisions to improve matching efficiency between content and ads. In terms of talent acquisition, Meta continues to recruit top professionals from leading industry institutions. Led by Yang Song, former head of TikTok’s recommendation system who joined Meta in 2025 advocating for reshaping recommendation systems through AI, alongside Lihong Li, an AI researcher formerly at Amazon. Additionally, researchers previously employed at OpenAI and Google have also been integrated into the broader MRS division. This move underscores Meta’s ongoing strategic commitment to strengthening its AI investments. Since launching the "Superintelligence Lab" (MSL) in summer 2025, the company has intensified efforts to build up AI talent reserves and explore leveraging AI to enhance ad performance. Industry analysts suggest this upgrade to the recommendation system may further solidify Meta’s competitive edge in social networking and advertising businesses.