USPS Warns of AI-Powered Scams
The U.S. Postal Inspection Service is warning that criminals are using artificial intelligence to make traditional scams more convincing. The alert, issued for National Consumer Protection Week, highlights how AI can create more realistic phishing emails, voice-cloning calls, and other fraudulent communications.
AI-powered phishing attempts are escalating, with some reports in 2025 indicating that up to 83% of these malicious emails were generated by AI. These tools allow criminals to create highly personalized and context-aware messages that are more effective at bypassing traditional security filters. This has contributed to significant financial losses, with business email compromise scams reaching $2.7 billion in annual damages. Voice cloning technology now enables scammers to replicate a person's voice with startling accuracy from just a few seconds of audio, often sourced from social media posts. This has supercharged "grandparent scams," where a cloned voice of a grandchild pleads for emergency funds. In one 2025 case, a woman lost $15,000 after receiving a frantic, AI-generated call from her "daughter." The financial impact of AI-driven fraud is substantial, with consumers reporting $12.5 billion in losses in 2024 alone, a 25% increase from the previous year. Investment scams accounted for the largest portion of these losses, totaling $5.7 billion. Some estimates suggest that over half of all fraud incidents now involve some form of AI. In one of the most audacious examples of this technology's power, scammers used a deepfake video to impersonate a company's chief financial officer in a video conference call. The hyper-realistic impersonation successfully convinced an employee to authorize a transfer of over $25 million. Beyond phishing and vishing (voice phishing), criminals are using AI to create synthetic identities by blending real and fabricated personal data. This tactic is fueling a rise in fraudulent loan applications and account takeovers, becoming the fastest-growing type of financial crime. In response, a new industry of AI-powered detection tools is emerging to counter these threats. Companies like Pindrop and CloudSEK are developing technology that analyzes audio for acoustic signatures and video for pixel-level inconsistencies to identify synthetic media in real-time. Federal agencies, including the FBI, have issued multiple warnings about these sophisticated scams. They advise the public to verify urgent requests independently and to be wary of any communication that pressures them into immediate financial action, even if the voice or video appears familiar.
