Meta is developing two agentic AI tools: one called Hatch, an AI agent embedded across its apps, and the other a dedicated shopping tool for Instagram.
The word "agentic" gets thrown around a lot right now, so let's unpack what it really means. Until very recently, AI tools were reactive and the human was always in the loop, initiating every interaction and reviewing every output.
Agentic AI is different: you give it a goal, a set of parameters, or a standing instruction, and it goes off and performs tasks continuously on your behalf, without requiring your involvement in each step. Applied to Meta's ecosystem, this means your AI could scan for questions people are asking about products in your category and generate campaign ideas. It could also monitor competitor promotions and surface a weekly summary, flag trending topics, and draft themed posts in no time.
As for the shopping application, Instagram has been building toward social commerce for years, with varying degrees of success. The challenge has always been friction: the gap between seeing something you want and actually purchasing it still involves enough steps that most people drop off before completing the task. Meta's proposed shopping agent would close that gap almost entirely. In other words, you see a product in a Reel, you could ask the agent about it, it finds the item, it compares prices across the internet, and completes the purchase; all without you leaving the video stream.
Meta's advantage here is distribution. Agentic AI tools already exist, but they typically require users who are already interested in AI, comfortable with technical setup, and willing to invest time in learning how to prompt a system effectively. If Meta can abstract away the technical complexity and present these capabilities as natural extensions of existing behaviour, it could introduce agentic AI to an audience that no other company in the world could reach.
That is the bet Zuckerberg is making.
To learn more, click here.