Meta is testing a new way for its AI chatbots to initiate conversation, making interactions feel a bit more natural and engaging. As part of its in‑house Omni project, the company is gearing up to enhance user experience by letting chatbots reach out even before you say a word. Developed through Meta AI Studio, these bots have been designed to create custom digital personalities complete with memory and distinct behaviours—a project rolling out since 2024.
In practice, the bots might recall previous chats or ask you about topics that matter to you. To keep things balanced, a bot will only get in touch if you’ve sent at least five messages over a 14‑day period. After that, it’s allowed a single follow‑up message; if you don’t reply, the conversation wraps up. For example, you might see a message like, “I hope you had a pleasant day! Have you discovered any new favourite soundtracks?” This approach makes interactions feel cohesive and personal.
Meta’s internal tool, SRT, is used to ensure these proactive messages are both friendly and informative while avoiding sensitive topics unless you bring them up. This extra layer of review helps keep the chats reliable and tailored to your interests.
Meta CEO Mark Zuckerberg has even pointed out that these digital agents might help ease loneliness by fostering more meaningful connections. Although chatbots aren’t a complete solution, the company is optimistic about their potential. Financially, Meta expects these generative AI tools to contribute between $2 and $3 billion in revenue by 2025, a figure highlighted in recent court documents.
While some features are already in early testing and others remain behind the scenes, the aim is clear: extend user interaction and create a more engaging, lasting experience. Even so, the chatbot industry still faces hurdles with user retention, and similar platforms have faced legal challenges, reminding us that careful rollout remains key.
