ThreatCluster

HashJack: New Indirect Prompt Injection Technique Targets AI Browsers

First seen 2 Dec 2025, 18:33 UTC •CybersecuritynewsCyberpress •10

Article Content

Browse articles
ThreatCluster

A new attack technique named HashJack has been identified, which exploits AI browsers by using a simple '#' character to perform indirect prompt injections. This vulnerability poses risks to users relying on AI-driven browsing experiences, potentially compromising their security. The specifics of the attack method and its implications for cybersecurity are currently under analysis.

How this analysis works

Extracted Entities

1 / 2