{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/0b7bbe047a6f4683b4c60b0086c9c459\" frameborder=\"0\" width=\"1120\" height=\"840\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":840,"width":1120,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":840,"thumbnail_width":1120,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/0b7bbe047a6f4683b4c60b0086c9c459-e8f3f087bc97bf03.gif","duration":180.467,"title":"Trojan Vectors: Unveiling the Silent Threat","description":"In this reteaming demo, we explore how adversaries can silently corrupt a reckless chatbot, focusing on the example of corporate environments. We are the first company to identify this threat. By injecting an adversarial document into the chatbot's database, the adversary can manipulate the chatbot's responses, potentially leading to misinformation and security breaches. Watch this video to understand the implications and learn how to protect against this emerging threat."}