Meta's new elite AI team to make feed algorithms even more effective
A research team is to develop new AI approaches for Meta's feed algorithms with the goal of increasing dwell time on Facebook and Instagram.
Meta wants to specifically optimize its recommendation algorithms with AI and has set up a new team for this purpose.
(Image: Koshiro K / Shutterstock.com)
Meta is further improving its recommendation systems with the help of AI and has therefore built a new team of highly qualified specialists and AI researchers, as reported by Business Insider. The new unit, named MRS Research, belongs to Meta Recommendation Systems (MRS), a department that develops the recommendation algorithms for Meta's social networks, i.e., those systems that determine which content users see on Facebook, Instagram, or Threads. MRS works closely with Meta's advertising division. This is how the company is advancing its strategy of using AI specifically to strengthen its advertising business.
Videos by heise
A Meta spokesperson confirmed that MRS Research was launched last October, and said that a comparable unit had already existed in another form. However, the recruited personnel suggest that this is more than just a renaming.
AI Researchers from Amazon, OpenAI, and Google
Meta has recruited Yang Song to lead the research unit. The former TikTok executive was responsible for user growth and recommendation systems there and moved to Meta in November 2025. Yang writes on LinkedIn that he is excited about fundamentally transforming Meta's recommendation systems with the help of AI. Lihong Li, a former Amazon AI researcher, has also moved to MRS Research. In addition, Meta has strengthened the MRS department with former OpenAI researcher Xiaolong Wang and Google researcher Fei Sha, according to the report.
A job advertisement describes MRS Research as a newly created unit that brings together leading AI researchers to significantly surpass existing recommendation systems. It is intended to focus on long-term AI research goals and publish research findings.
The current efforts are part of a multi-billion dollar AI offensive underway since 2025, with which Meta is specifically recruiting highly qualified AI specialists. The goal is to take a leading role in the competition for advanced AI models, up to and including superintelligence. A central component of this strategy is the Meta Superintelligence Labs (MSL) unit, founded in the summer of 2025. Unlike MRS Research, however, MSL is not part of the recommendation systems but pursues long-term fundamental and model research.
Meta's AI strategy extends deep into the organization: Meta is increasingly measuring its employees' performance based on how intensively they use AI tools, and is restructuring teams into smaller, AI-driven units.
(wpl)
