A New York startup called Asymptote Labs has raised $5.5M to help companies see what their AI agents are doing, and stop them when an action looks risky. Tower Research Ventures led the seed round, the company said on Tuesday. Lytical Ventures, 1984 Ventures and South Park Commons also took part.
What the tools do
AI agents are programs that take actions on their own, such as running commands or editing files. Asymptote's free tool, Beacon, records that activity. It works with Claude Code, Codex, Cursor and more than 20 other agent tools. Prompts, commands and file edits go into one standard format, so security teams can see what each agent did and on whose behalf.
Beacon is open source, which means its code is public and free to use. The paid platform builds on it and can block risky actions, such as data leaks, before they happen, the company said. The money will go toward engineering, hiring and sales.
What is not known
Asymptote did not disclose its valuation. It said security teams already use its tools in production but did not name any customers. Tower's own security team is testing Beacon on its computers. Beacon launched in May, and the company says its GitHub page grew its stars by more than 460% in September. Those numbers show interest, not how well the tool works in practice.
The takeawayIf AI agents can touch your files or customer data, ask who can see what they did and who can stop them.