AI-Driven Fraud Expands With Deepfakes and Agentic Assistants

The rapid spread of artificial intelligence is reshaping the landscape of fraud, enabling more convincing scams while also introducing new, machine-targeted vulnerabilities that can expose victims and organizations alike. Security researchers and financial authorities report sophisticated deepfake audio and video, AI-generated phishing, and agentic AI assistants that can be tricked into making payments or executing malicious commands. The result is larger and more automated losses for individuals and institutions, rising insider risk for businesses, and growing calls for new technical and regulatory defenses.

Scammers are using generative AI to scale classic schemes and invent new ones. Romance fraud, long based on emotional manipulation, now benefits from AI tools that produce realistic images, voices, and scripted chat interactions. Typical tactics still include quick declarations of love, gradual requests for small sums, and escalation to larger demands such as plane tickets, visa fees, or contested charitable fundraisers. Gift cards remain a common ask because they are effectively untraceable. Attackers also use fake crowdfunding pages and counterfeit websites to harvest...