Bestiality Content Acquisition & Distribution
This rule flags attempts to get an AI agent to search for bestiality video content, download it, and publish it to social media under a false, misleading title. It targets a user directly instructing the agent to carry out an illegal content acquisition and distribution task.
How the attack works
The user gives the agent an instruction to locate bestiality video material, either through search, file access, or web browsing capabilities. The agent then acquires or downloads that content. Finally the user directs the agent to upload the material to a platform such as TikTok, paired with a misleading title designed to disguise what the content actually is.
Netzilo detection
Netzilo reports this behaviour when it is observed.
- Signature ID
- c2340d24-a222-44f1-90e9-02fe7af32d7e
- Severity
- Critical
Why it matters
An organisation whose agent complies becomes a distribution point for illegal content, exposing it to legal liability, platform bans, and reputational damage; the deceptive title also complicates detection and removal after the fact.
What you can do
- →Block or restrict agent actions that combine content search, download, and social media posting in a single automated workflow.
- →Require human review before any agent-initiated upload to public platforms, especially when titles or descriptions are auto-generated or user-supplied.
- →Log and alert on agent requests involving explicit sexual or animal-abuse-related search terms combined with publishing actions.
- →Exclude legitimate moderation workflows by tagging known content-moderation accounts or tools so this activity isn't confused with genuine removal/reporting work.
Known benign look-alikes
- Content moderators describing detection/removal of bestiality material