Meta is assembling an elite new AI lab for its recommendations division

Meta is embarking on an ambitious initiative to assemble a team of top-tier AI researchers, aimed at refining the algorithms that keep users engaged on platforms like Facebook and Instagram. This new group, known as MRS Research, is a component of Meta's Recommendation Systems, which manages the content feeds across its applications. The recruitment strategy has already attracted professionals from major players like TikTok and Amazon. The MRS division's restructuring in October aimed to consolidate various teams under one umbrella, according to a spokesperson. While the group has evolved forms over recent years, the current focus of MRS Research is to pursue long-term AI research objectives and publish innovative studies that will enhance Meta's recommendation engine. Described as a 'newly created organization', MRS Research is set to unite exceptional AI researchers and engineers with the goal of significantly advancing existing AI recommendation systems. This latest development underscores Meta's commitment to leveraging AI to bolster its core revenue streams, particularly advertising. Yang Song, who oversees this initiative as Meta's vice president of recommendation research, expressed enthusiasm on LinkedIn about the potential to 'revolutionize' the company’s recommendation systems with AI. He joined Meta in November 2025 after leading user growth and recommendation strategies at TikTok. Following his arrival, the team also welcomed Amazon AI researcher Lihong Li, along with other notable experts like Xiaolong Wang from OpenAI and Fei Sha from Google. Although Meta declined to provide comments on specific personnel within the team, the company has been aggressively expanding its AI talent pool since the summer of 2025. This included the establishment of the new Meta Superintelligence Labs (MSL), led by former Scale AI CEO Alexandr Wang. Notably, MRS operates independently of MSL. In addition to these efforts, Meta has recently brought on board three AI researchers from the notable startup Thinking Machines Lab, further solidifying its commitment to AI advancements. The company is rebranding its workforce as 'AI builders' and encouraging employees to integrate AI into new product development. Meta has been actively exploring innovative AI technologies to enhance the algorithms that power its profitable applications, including a recent AI model that reportedly improves ad performance by delivering more relevant advertisements to users.