iLands’ user-made AI agents surge past 90,000 as the experimental platform goes global

Recent coverage abroad highlights how fast “iLands,” a so‑called “user-generated agent network,” is growing just months after launch. According to a summary on Wikipedia, the platform went live on July 27, 2026, and users had already created around 70,000 AI agents by September and more than 90,000 by October. Accessible via web and mobile apps, iLands lets users design autonomous AI personas—defining their traits and behavior rules—and then share them with others for chat, role‑play, or utility tasks.

The company behind iLands is a China‑based startup whose founder has claimed to be a former ByteDance employee, linking the project loosely to the culture that produced TikTok. On the platform you can find everything from casual chat companions and study helpers to game‑like characters with elaborate backstories. Some observers describe it as “a social network for AI agents,” since users can follow each other’s creations and interact inside a shared environment.

Key questions remain about language support, safety policies, and data usage—there is no clear guarantee yet that it handles privacy or misinformation robustly. Still, services like iLands point toward a future where ordinary users don’t just talk to AI, but actively publish and trade their own agents. For Japanese observers watching global AI and social trends, it’s a noteworthy experiment that could influence how next‑generation social platforms are designed.