Consumer vs Enterprise Browser Labor
David Mlcoch, co-founder & CEO of Asteroid, on browser automation and the last mile problem of AI
The key split is consumer browsing versus enterprise browser labor. Dia, Comet, and a possible OpenAI browser are built around helping one person browse, search, summarize, and complete occasional actions inside a local browser. Asteroid is building a hosted browser system that lets companies run thousands of repeatable workflows, monitor recordings, and turn back office data entry into centralized software operations.
-
Asteroid sits closer to Browserbase than to Dia or Comet on infrastructure. Both Asteroid and Browserbase are cloud browser products, but Browserbase is aimed at developers hosting and managing headless browsers, while Asteroid is packaging that capability for operations teams that need reliable, supervised task runs in insurance and healthcare.
-
The practical difference is scale and control. Consumer AI browsers run on an employee's device and help with a single research or form task. Asteroid runs hosted browser workers in parallel, then exposes live view, recordings, and supervision so an enterprise can treat browser work like a managed workforce instead of a desktop assistant.
-
The overlap is real because the underlying model capability is converging. Asteroid already uses OpenAI and Anthropic models for browser tasks, and broader platform players are shipping browser products with built in AI. That means the interface can look different while the competition moves toward the same end state, software that can act on the web for users.
Over time the browser market is likely to separate into a personal copilot layer and an enterprise execution layer. The winners in enterprise browser automation will be the companies that can make browser work cheap, auditable, and dependable enough to replace outsourced back office teams, not just make an individual user faster at browsing.