Meta Launches Public AI Bot on Threads in Limited Beta
Key Points
- Meta tests a public AI chatbot on Threads using the handle @meta.ai.
- Beta currently limited to users in Malaysia, Saudi Arabia, Mexico, Argentina and Singapore.
- Bot runs on Meta's Muse Spark model, also deployed on WhatsApp, Instagram, Facebook and Messenger.
- Replies are posted publicly in the same language as the original post.
- Users cannot block the bot, leading to a wave of criticism shortly after launch.
- Feature mirrors X's AskGrok but differs in visibility and lack of user controls.
- Meta has not announced plans to expand the test or add a block option.
Meta has begun testing a public-facing AI chatbot on its Threads platform, assigning the bot the handle @meta.ai. Users in five countries can tag the account in posts or replies and receive real‑time answers powered by Meta’s Muse Spark model. The feature mirrors X’s AskGrok, but unlike the X tool, the Threads bot posts its replies publicly and currently cannot be blocked, sparking a wave of criticism from early users.
Meta is rolling out a new artificial‑intelligence feature on Threads that lets users tag a dedicated bot account, @meta.ai, directly in their posts or replies. When mentioned, the bot replies publicly with context, recommendations, or factual information about the topic at hand. The functionality mirrors the widely used AskGrok on X, where tagging an AI has become a routine part of conversation for millions of users.
The beta is limited to five markets – Malaysia, Saudi Arabia, Mexico, Argentina and Singapore – and is being offered to a small group of participants. In those regions, anyone can write a question such as “Why is everyone talking about the Met Gala right now?” or “How are the Knicks doing in the playoffs?” and the bot will generate a response in the same language as the original post. The multilingual capability is designed to accommodate Threads’ global audience.
Behind the scenes, the bot runs on Meta’s latest Muse Spark model, the same large‑language model that powers AI features across WhatsApp, Instagram, Facebook and Messenger. On WhatsApp, Meta is also testing a private “side chats” version that supplies answers only to the user who asks, a stark contrast to the fully public nature of the Threads implementation.
Early reactions have been mixed. Within hours of the @meta.ai account going live, users discovered they could not block the bot, prompting a flood of angry posts on the platform. Critics point to the experience with X’s Grok, which drew backlash after generating non‑consensual sexualized images and led the company to add a user toggle to block the AI. While Meta has historically applied stricter guardrails to its models, the decision to make the bot’s replies visible to everyone raises concerns about misinformation, harassment and the potential for the tool to be weaponized in public discourse.
Meta has not disclosed a timeline for expanding the test beyond the current markets, nor has it indicated whether a blocking feature will be added. The company’s broader AI rollout strategy appears to be focused on integrating Muse Spark across its family of apps, offering both private and public interaction modes. Observers will be watching how Threads users adapt to a chatbot that answers in the open, and whether the platform can balance utility with the need for content safeguards.
Industry analysts note that the move underscores Meta’s ambition to make AI a core part of social interaction, positioning Threads as a testing ground for features that could eventually appear on its larger platforms. As the beta progresses, Meta’s handling of user feedback and moderation challenges will likely shape the future of AI‑driven conversation on social media.