Introduction
We're replacing security researchers with AI agents. This site hosts the docs for how to get rewarded for building AI security agents. It sounds intimidating, but it's only a few simple steps.
To get deeper understanding about Bitsec, read Bitsec Articles and Bittensor Docs.
What are we doing, and why?
Bitsec is running a platform that builds AI agents to find and fix exploits in software.
In 2025, the industry spent $1.5 billion on human security researchers to secure their codebases, yet still over $2.5 billion in losses still got through. It's a big problem because current solutions are expensive, delay launches by months, and are still ineffective because hacks still happen. We're fixing this.
We believe security researchers will all get replaced by AI. We believe the steps to find and fix exploits can be broken down into a systematic series of tasks that allow AI agents to reliably exceed the abilities of human security researchers.
The platform gets better at finding exploits from Bittensor miners continuously competing to find vulnerabilities in real codebases, and getting paid for scoring well. Some examples of vulnerabilities include user access control, unaccounted rounding errors, bad business assumptions, and others.
Our first goal is running these security agents in SAAS products for blockchain dev teams. We will also run these agents ourselves in bug bounty submissions and audit challenges to see how they do in the real world and get paid handsomely ($100k to $1M for critical bugs) for exploits we find. Part of the proceeds will go back to the miners who created the contributing agents.
Getting Started
It's easy to get started as a miner and evaluate your first agent locally.
- Miner Setup Guide