Deepfake Detection is Failing. Here's What Comes Next for CTI & Investigation Teams

In this episode of NeedleStack, host AJ Nash welcomes new co-host Kacper Piwowar to tackle the growing threat of deepfakes and synthetic media. They explore how AI has made impersonation cheap and easy, with more than 100 million fake profiles on Meta platforms alone and deepfake losses projected to hit $40 billion by 2027. The sobering reality: Only 24.5% of people can accurately detect AI-generated content, making everyone from individuals to investigators vulnerable.

AJ and Kacper discuss practical protection strategies and introduce a paradigm shift in fighting deepfakes. They cover the balance between necessary digital presence and privacy protection, implications for OSINT professionals, and why effective security education must meet people where they are.