Meta Quietly Rolls Out 'Manus' AI Ad Agent

Meta's acquisition of Manus, a Singapore-based startup, for a reported $2 billion is a significant move to embed autonomous AI agents directly into its advertising platform. The startup, founded in China in 2022 and later relocated, had already developed a functioning product with a growing user base before the acquisition. The rapid integration of Manus into Ads Manager, within approximately seven weeks of the deal closing in December 2025, underscores Meta's push to generate measurable returns from its substantial AI investments. The core of Manus is its general-purpose autonomous AI agent, designed to execute multi-step tasks with minimal human supervision. Unlike traditional chatbots that respond to queries, Manus is given a goal which it then breaks down into a sequence of sub-tasks, selects the necessary tools, and executes them to deliver a completed result. This allows the agent to handle intermediate decisions itself, moving beyond simple automation to a more "agentic" workflow where the AI plans and executes complex campaign elements. This integration is part of a broader trend of "agentic workflows" in marketing, where AI agents make real-time decisions within a customer journey, adapting to changing behaviors. These systems are designed to operate autonomously, learn continuously, and collaborate to achieve goals, shifting the paradigm from rigid "if-this-then-that" automation to dynamic, goal-oriented execution. This allows marketers to move from being task-doers to strategic thinkers, with AI handling the execution and optimization. The introduction of Manus builds upon Meta's existing suite of AI-powered advertising tools, known as Advantage+, which already automates aspects like budget allocation, audience targeting, and creative optimization. Meta's vision is a "goal-only" ad system where marketers input an objective and budget, and the AI manages the entire campaign lifecycle. This is powered by technologies like Meta Lattice, a machine learning architecture trained on trillions of ad signals to improve audience discovery and conversion-focused delivery. The move toward agent-driven media buying is not unique to Meta, with the broader ad industry preparing for a future where AI agents control significant portions of media budgets. Initiatives like the Ad Context Protocol (AdCP) are emerging to create standardized communication methods for AI agents across different platforms, aiming to prevent fragmentation and ensure interoperability. This reflects a growing consensus that agentic AI will become the default logic layer for planning, activation, and optimization in programmatic advertising.