DepthFirst: “We found more bugs that Mythos missed, at one-tenth the cost”

Cybersecurity startup DepthFirst says its AI model found more bugs than Anthropic’s Mythos at about one-tenth the cost.

The company also launched an Open Defense initiative, offering $5 million in credits for security testing.

DepthFirst says it has already uncovered vulnerabilities in NGINX, Linux, Google Chrome, and FFmpeg.

The claims highlight how AI security tools could improve vulnerability discovery while cutting costs for open-source and internet infrastructure.