The vulnerability in TypeBot’s OpenAI “Create Transcription” action allows the server to retrieve audio from any URL supplied by a user. Because the handler omits the SSRF protection that is present in other parts of the code, an attacker can trick the server into making HTTP requests to arbitrary internal or localhost addresses. The retrieved content is forwarded to OpenAI’s Whisper API and the resulting transcription is returned to the attacker, providing a covert channel for internal information exposure. This flaw represents a classic SSRF weakness (CWE‑918).

TypeBot, developed by baptisteArno under the typebot.io brand, is affected in all releases prior to version 3.16.0. Users running any legacy version of the chatbot builder should verify their installed version and plan an upgrade if necessary.

The CVSS score of 5.4 indicates moderate severity, with an EPSS score of less than 1% suggesting a low but non‑zero likelihood of exploitation. The vulnerability is not listed in CISA’s KEV catalog. Attackers could exploit the flaw by supplying a malicious URL to the transcription feature, causing the server to access internal resources. Because the flaw is input‑validation based and does not require elevated privileges, the risk is largely confined to the server’s network context, but could facilitate further lateral movement if the internal services expose sensitive data.