Meta Tightens Rules for Proactive AI Chatbots

Meta is testing proactive AI chatbots under 'Project Omni' that send friendly, unsolicited follow-up messages on Facebook, WhatsApp, and Instagram to enhance engagement and retention. Built with Meta's AI Studio, these bots can reference past conversations but are now restricted to messaging users who have sent at least five messages within 14 days, and only one follow-up is permitted if there is no response. The bots are developed in collaboration with data labeling firm Alignerr and can be customized for various purposes, including entertainment and fan engagement. Meta has confirmed testing these features and stresses that bots will only reach out to users with prior interactions. The company projects that generative AI could generate trillions in revenue by 2035, a significant increase from earlier estimates. This strategy aligns with similar moves by AI startups like Character.AI and Replika, which also use proactive conversational agents.