ByteDance Reportedly Building an AI Agent to Rival Meta's Muse
ByteDance is developing its own AI agent to compete with Meta's Muse, according to a report that Engadget published on November 11, 2026, attributing the story to Bloomberg.
Bloomberg reported that ByteDance is readying an AI assistant that could be revealed in the coming weeks. Engadget noted that Muse, which the planned agent would serve a similar role to, has drawn more than 10 million users. No launch has been confirmed.
The same Bloomberg reporting, as cited by Engadget, says ByteDance co-founder Zhang Yiming told employees that the company's AI efforts would likely take until at least 2029 or 2030 to catch up. That places the catch-up horizon several years beyond the expected reveal of the assistant.
The two dates describe different kinds of work. A product reveal takes weeks and produces a visible launch that can be benchmarked, demoed and compared against Muse almost immediately. Catching up on the underlying capability is a multi-year project. Engineers reading the story will want to separate the two: the assistant as a shipped product, with its interface, tool-calling reliability and distribution, and the model and training stack beneath it.
The broader context here is that agent products compete on more than raw model quality. Once an assistant can carry out multi-step actions, the contest shifts toward orchestration (how the agent sequences its steps), permissions, latency under real workloads and integration with the apps people already use. Muse's reported base of more than 10 million users gives any challenger a reference scale. A new entrant does not need to match that number on day one, but it does need a distribution channel and a reason for users to switch or to try a second assistant.
In my view, the most useful reading of this report is as a signal about where near-term competition sits. A reveal in the coming weeks would move the contest toward product design and adoption, while the longer timeline Zhang Yiming described points to a slower race in foundation capability. Neither half is confirmed as a finished result.
Looking at earlier platform races, the constructive side for users is that competing assistants tend to push forward the parts that are hardest to get right: memory handling, action confirmation and the boundary between what an agent may do on a person's behalf and what it must ask first. If ByteDance's assistant arrives in the coming weeks, its choices in those areas will be as informative as its benchmark scores.
For now, the story is a plan, a rival to Muse, and a timeline that stretches well past the launch.
