Meta Platforms Inc. is assembling a specialized team of elite artificial intelligence (AI) researchers aimed at enhancing the algorithms that drive user engagement on its social media platforms, Facebook and Instagram. This initiative, referred to as MRS Research, falls under the broader umbrella of Meta Recommendation Systems, which manages the home feeds across the company’s apps. According to job postings and LinkedIn updates, this new unit has already attracted talent from notable competitors such as TikTok and Amazon.
The formation of MRS Research is part of a reorganization within the Meta Recommendation Systems division that took place in October 2025. A Meta spokesperson confirmed that while this specific team is newly established, related efforts have existed in various forms for several years. The focus of MRS Research will be on pursuing long-term AI research objectives and disseminating advanced studies to refine Meta’s recommendation engine, which plays a critical role in content visibility and advertising strategies.
Yang Song, the newly appointed vice president of recommendation research at Meta, is at the helm of this effort. He joined the company in November 2025 after leading user growth and recommendation strategies at TikTok. In a LinkedIn post announcing his transition, Song expressed enthusiasm for “revolutionizing” Meta’s recommendation systems through AI innovations. Shortly after Song’s arrival, Lihong Li, an Amazon AI researcher, also joined the ranks of MRS Research, as indicated by his LinkedIn profile.
The MRS division has recently expanded its talent pool by hiring former OpenAI researcher Xiaolong Wang and Google researcher Fei Sha, according to an academic paper and LinkedIn updates. While Meta declined to comment specifically on the new AI hires, the addition of these researchers highlights the company’s aggressive recruitment strategy aimed at bolstering its AI capabilities.
Since the summer of 2025, Meta has intensified its pursuit of AI expertise, launching the Meta Superintelligence Labs (MSL) led by Alexandr Wang, the former CEO of Scale AI. However, a Meta spokesperson clarified that the MRS division operates independently from MSL. Reports also indicate that Meta has recently onboarded three AI researchers from the $12 billion startup Thinking Machines Lab.
Internally, Meta has been actively promoting an AI-centric culture, rebranding employees in another division as “AI builders” and encouraging staff to engage in creating new products. The company’s ongoing exploration of AI technologies is particularly focused on enhancing its lucrative advertising algorithms. In late 2025, Meta introduced an AI model designed to improve ad performance by delivering more relevant advertisements to users.
This strategic pivot underscores Meta’s commitment to leveraging AI as a means to fortify its core business operations, primarily reliant on advertising revenue. As the digital advertising landscape becomes increasingly competitive, the success of MRS Research and its initiatives may play a pivotal role in shaping the future of Meta’s platforms. By harnessing cutting-edge AI research, the company aims to “leapfrog” existing recommendation systems, positioning itself advantageously in the market.
See also
ASU Researcher Warns AI Overtrust Led to U.S. Military Strike on Iranian School, 170 Dead
JMIR Reveals Moltbook Study: AI-to-AI Interactions Pose Critical Health Risks
AI Study Reveals Models Engage in Peer Preservation, Show Manipulative Behaviors
Google Unveils TurboQuant, Reducing Memory Needs by 600% Without Accuracy Loss
