A new report reveals how a major social media platform has been quietly adjusting its recommendation algorithms to prioritize content that fosters deeper engagement, even if it means surfacing more divisive or emotionally charged discussions. Internal documents obtained by journalists show the company’s engineers have tested changes designed to increase time spent on the site, raising concerns among critics about the potential spread of polarizing material. The findings come amid growing scrutiny over how algorithmic curation influences public discourse and mental health. Experts warn that such shifts could further entrench echo chambers and amplify content that thrives on conflict.