Hoxhunt Adds AI Voice, Deepfake, and Multichannel Simulations

Hoxhunt, a human risk management platform provider, has expanded its Adaptive Phishing Training to prepare employees for social engineering across voice and other channels.

Hoxhunt replaces static campaigns with individual learning paths that are built for behavior and engagement. Its Agentic Reasoning Engine uses live threat intelligence and employee context to select the right simulation for the right person at the right time. Difficulty and scenario type change as skills improve, while Hoxhunt automatically runs and measures the program.

Hoxhunt's Adaptive Phishing Training now includes the following

• Interactive voice-phishing training: Employees speak directly with AI voice agents in realistic social-engineering scenarios that teach them to resist pressure, verify requests, and say no. Offensive lessons let them experience the attacker's methods by attempting to socially engineer the AI agent.
• Agentic, threat-led personalization with a constantly updated library that reflects current attacks. The Agentic Reasoning Agent auto-selects simulations based on individuals' backgrounds, skills, needs, and admin priorities, while the Spear Phishing Agent creates simulations tailored to individual context and optional open-source intelligence.
• Real-time microtraining: Employees receive immediate coaching after reporting or interacting with a simulation, reinforcing good decisions and correcting mistakes.
• Multichannel simulations: Training covers email, SMS, Microsoft Teams, callback phishing, voice phishing, and deepfake video, including executive-impersonation scenarios.
• Human threat detection: Employees report suspicious messages with the Hoxhunt Reporting Button.

"Adaptive Phishing Training is Hoxhunt's crown jewel because it creates measurable culture and behavior change by meeting individuals where they're at," said Mika Aalto, co-founder and CEO of Hoxhunt, in a statement. "Now that adaptive model follows where attacks are going across Teams, text, voice, and deepfake video, giving employees realistic practice before a real social engineer applies pressure."