Researcher Uses AI to Chain Exploit and Reach PostHog's Production Database

A researcher testing analytics provider PostHog discovered that its heatmap screenshot feature launched the Chromium browser using the '--no-sandbox' flag, a configuration that removes an important security barrier normally built into the browser. Rather than simply reporting the misconfiguration, the researcher worked with an AI coding assistant to build a proof-of-concept exploit targeting a known Chromium vulnerability affecting the version PostHog was running.

After roughly 36 hours of iterative work, including switching between different versions of the AI assistant, the exploit successfully achieved remote code execution inside a PostHog production worker container that had access to the company's production database. The researcher noted that a newer AI model version performed the task more reliably and with fewer shortcuts than an earlier one, highlighting how quickly AI tools are becoming capable of assisting in complex exploit development.

This case shows that browser automation features, such as those used for generating screenshots or thumbnails, can become serious attack surfaces if not properly isolated. It also demonstrates that AI assistants can now meaningfully speed up the discovery and weaponisation of known vulnerabilities, raising the stakes for organisations that rely on browser rendering in backend services.