Meta is enhancing its AI chatbot with new capabilities designed to move it beyond simple question-answering and image generation. The company announced on July 24, 2026, that Meta AI will now integrate with users' calendars, offer automated daily briefings, and perform steerable research. This update signifies Meta's intensified effort to compete with leading AI assistants such as ChatGPT, Gemini, and Claude.
The revamped Meta AI can now access a user's calendar to help plan events and provide daily summaries. According to Meta, users can set up recurring tasks, like weekly meal plans or trend updates, which the AI will then manage without further prompting. This marks a shift from an AI that answers questions to one that takes action. The assistant is also capable of conducting in-depth research, synthesizing information from the open web and across Meta's own applications. Users can guide this research process, directing changes in tone or focus as the AI generates its output. All AI-generated content is now consolidated in a single location within the app for review and sharing.
These new functionalities are powered by Muse Spark 1.1, an AI model released by Meta's Superintelligence Labs on July 9, 2026. Meta describes Muse Spark 1.1 as an agentic model built for planning and tool use, featuring a one-million-token context window. It can orchestrate multiple sub-agents and operate directly on a computer, capable of writing scripts or navigating interfaces. The company stated that the model underwent testing against frontier-risk categories prior to its release.
Meta's approach to AI assistants has evolved with this update. Previously, the company had suggested it would not pursue productivity tools that directly copied rivals. However, this new direction directly challenges established players in the AI assistant market. Meta frames these advancements as progress toward "personal superintelligence," a concept CEO Mark Zuckerberg has highlighted as the future of AI.
The rollout of these features is occurring in select markets via the Meta AI app and meta.ai. Meta plans to extend availability to additional countries and platforms, including WhatsApp, in the coming weeks. The company has stated that Meta AI will remain free, contrasting with subscription-based models offered by some competitors. Meta also emphasized that users retain control over their interactions and that incognito chats remain available for private conversations.
Examples of Meta AI's new capabilities include assisting with home renovation projects by searching for furniture within a specified budget and creating mood boards, or building a personalized half-marathon training schedule. These demonstrations illustrate the AI's potential to manage complex, multi-step tasks.
Meta's prior work in AI includes research in natural language processing, with initiatives like the Seamless models for real-time voice translation and efforts to develop machine translation for less common languages. The company also developed PyTorch, an open-source machine learning framework. While Meta AI has historically focused on research, this latest update signals a stronger push into consumer-facing AI applications designed for daily productivity.