AI Impersonation of Marco Rubio: State Department Investigates Cyber Operation Targeting Diplomats

The U.S. State Department is investigating a cyber operation involving artificial intelligence (AI)-generated impersonations of Secretary of State Marco Rubio, according to an internal cable obtained by multiple news outlets. Between July 3 and 8, 2025, attackers used AI voice cloning and fabricated credentials (e.g., “[email protected]”) to send text messages and voice calls to foreign diplomats and U.S. officials via encrypted platforms like Signal^{1, 2}. The FBI has attributed the activity to a Russia-linked threat actor³.

Incident Overview

The impersonation campaign targeted three foreign ministers, a U.S. governor, and a member of Congress⁴. Attackers replicated Rubio’s vocal patterns using AI tools, leaving voicemails and sending messages that appeared to originate from State Department channels. The July 3 cable described the attempts as “unsophisticated but credible,” noting no direct compromise of State Department systems but warning of potential information leakage⁵. This follows a May 2025 incident where AI was used to impersonate White House Chief of Staff Susie Wiles³.

Technical Analysis

The operation exploited gaps in authentication protocols for high-level communications. Attackers bypassed multi-factor authentication (MFA) by using AI-generated voice calls to manipulate targets into sharing sensitive data⁶. The State Department has since mandated MFA for all diplomatic channels and partnered with the FBI to trace the activity to known Russian cyber infrastructure⁷.

Tool/Method	Description	Source
AI Voice Cloning	Used OpenAI’s Voice Engine or equivalent to mimic Rubio’s speech	LiveNOW Fox
Signal Spoofing	Fabricated sender IDs to appear as State Department accounts	BBC
Credential Harvesting	Spear-phishing links embedded in messages	PBS

Response and Mitigation

The State Department’s Cybersecurity Directorate issued updated guidelines for verifying high-priority communications, including:

Mandating secondary verification via secure channels for sensitive requests
Deploying AI-content watermarking for official correspondence
Restricting Signal use to pre-vetted contacts

David Axelrod, former Obama adviser, noted the broader implications:

“This is the new world we live in. We’d better figure out how to defend against it”⁵.

Policy and Historical Context

No federal laws currently prohibit AI voice cloning for impersonation¹⁰. The incident mirrors the 2024 fake Biden robocalls in New Hampshire, highlighting the escalating use of AI in political cyber operations. Cybersecurity firms advocate for international standards on AI-generated content verification¹¹.

Conclusion

The Rubio impersonation campaign underscores the evolving threat of AI-driven social engineering. Organizations should prioritize voiceprint authentication and staff training to detect synthetic media. The lack of regulatory frameworks for AI misuse remains a critical vulnerability in diplomatic and corporate security postures.