This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
The FBI issued a public service announcement regarding the use of deepfake video and audio technology by malicious actors to impersonate job applicants during remote work interviews. By using stolen PII and deepfakes, these actors successfully applied for roles that granted them access to sensitive corporate data, including financial information and customer PII. The report highlights the lack of synchronization between visual and auditory cues as a key indicator of the deepfake usage.
Voice and video deepfakes were reported by FBI Internet Crime Complaint Center (IC3) in complaint reports to have been deployed during online interviews of the candidates for remote-work positions.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to gain a personal advantage over others such as through cheating, fraud, scams, blackmail or targeted manipulation of beliefs or behavior. Examples include AI-facilitated plagiarism for research or education, impersonating a trusted or fake individual for illegitimate financial benefit, or creating humiliating or sexual imagery.
Also classified as
The fraudulent applications relied on stolen personally identifiable information (PII) of real individuals to bypass background checks.
The incident involves malicious actors using synthetic media (deepfakes) to impersonate job applicants for fraudulent employment and unauthorized access to corporate systems, which constitutes fraud and targeted manipulation.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by human malicious actors who intentionally utilized post-deployment deepfake technology and stolen PII to deceive employers during remote job interviews.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
FBI warning highlights malicious actors using deepfakes and stolen PII to secure remote IT and programming positions, gaining unauthorized access to corporate databases, financial data, and proprietary information. While representing an innovative fraud vector, the national security impact remains minor and manageable through standard corporate security and identity verification protocols.
The deepfake video and voice generation systems were used as tools by human malicious actors to generate specific synthetic media outputs based on direct inputs, without exhibiting autonomous planning or agency.
Risk Level 3. Limited Risk: The report describes the use of AI-generated content such as deepfakes, which are subject to transparency obligations under the Act to ensure users are aware they are interacting with synthetic media.
The AI systems were used to generate synthetic video and voice to impersonate job applicants during remote interviews.
The FBI PSA provides clear, authoritative details on the methodology of the fraud, specifically identifying the use of deepfakes and stolen PII. While specific statistics on victims and financial losses are not quantified, the nature of the AI safety risk and its classification under the taxonomies is highly unambiguous.
The incident could be viewed primarily as a traditional cybersecurity/identity theft issue, but the central role of synthetic media (deepfakes) to bypass human verification makes it an AI safety incident.
Technology
Voice Spoofing
Voice deepfakes are used to spoof applicant audio during live interviews.
Deepfake Video Manipulation
Altered video misrepresents the applicant's identity and live actions.
Data Inputs
Stolen PII
Stolen personal data is used to bypass initial identity verification.
Human Factors
Delayed Detection by Interviewer
Interviewers fail to immediately spot audio-visual synchronization issues.
Process and Methods
Remote-Only Interviewing
Virtual interview processes lack physical verification of applicant identity.
Insufficient Background Checks
Pre-employment checks fail to detect stolen identities early in the process.
AI system involved
An AI system was involved, but the reports do not identify it.
Identified by automated extraction; can differ from the alleged developers above.
