Beating Deepfakes and AI Threats: Jason Thatcher on Cybersecurity’s New Frontier
Cybersecurity | AI Threats | Social Engineering

Introduction
Are your eyes and ears enough to keep your business safe? Jason Thatcher, cybersecurity expert and founder of breacher.ai, joins the Multiverse Technology Podcast to sound the alarm on the next generation of digital threats: AI-powered phishing, deepfakes, and intelligent social engineering. This is the new arms race and everyone is a target.

The Rise of AI-Powered Threats
After decades of defending against basic phishing emails, companies now face adversaries armed with AI voice and video cloning, agentic bots, and other advanced tools. These aren’t coming they’re here, and attacks are multiplying. Jason shares stories of C-suite executives being impersonated by deepfakes, and how traditional defenses are no longer enough.

Trust Is Under Attack
The new attacks target the “human trust layer,” not just technology or process. Even sophisticated cybersecurity stacks can’t stop a convincing deepfake phone call or video impersonation. As organizations get bigger and more complex, the risk skyrockets especially when employees are scattered and less likely to recognize each other’s voices or faces.

The STOP Framework for Staying Safe
Jason shares his four-part STOP framework for defense:

• Slow down:Don’t act under pressure pause, breathe, and verify.

• Trust less:Take nothing at face value; always question requests for information or action.

• Origin and out-of-band verification:Always use a second channel or different method to confirm any critical request.

• Policy and process:Keep corporate guidelines updated and rigorously tested; train everyone, not just IT.

Training and Testing the Human Layer
True security, Jason argues, depends on immersive, role-specific, scenario-based training. Using deepfake bots and simulations, his team tests and teaches organizations to recognize, resist, and respond to AI-driven social engineering. Leadership, employee buy-in, and relentless curiosity are vital not just technical fixes.

Key Takeaways

• AI deepfakes and agentic bots have made traditional security measures obsolete

• Human psychology specifically authority bias remains a primary weakness

• Testing and immersive training are the only way to ensure policies work in real scenarios

• Every employee, not just IT, must be empowered to act as a “human firewall”

• In cybersecurity, trust is now a vulnerability education is the best defense