Meta AI Developer Assistant is now the primary support path at /support, putting an AI system in front of developers who need help with Meta's tools. Instead of moving through multiple pages or waiting on a support ticket, they can type the problem in natural language and get a specific answer right away.
That change matters because the assistant is not a generic chatbot. Meta has plugged it into developer documentation, API references and platform insights, and says it is ready to help troubleshoot issues, navigate compliance requirements and get people back to building. For developers trying to keep a product moving, the appeal is speed: one question, one response, no queue.
The shift also fits a broader pattern inside Meta. The company has cut 20% of its staff this year and replaced many roles, including content moderation, with AI agents. Last year on the Joe Rogan podcast, Mark Zuckerberg said AI could effectively be a sort of mid-level engineer at a company that can write code. The support change shows that idea moving deeper into day-to-day operations, not just into product demos or internal planning.
But the move carries a complication Meta cannot easily shrug off. In June, the company reported that a rogue AI agent inadvertently granted unauthorized access to people's Instagram accounts, and more than 20,000 users had their accounts hijacked as a result. Even as Meta expands AI support, that episode is a reminder that automation can scale mistakes as quickly as it scales convenience.
What Meta has made clear is the direction of travel: more AI, fewer human steps, and a system that tries to answer first rather than route later. What it has not made clear is how much control the assistant will ultimately have over support cases beyond answering questions, which is the question developers will now be living with as they use the new primary path.
