Reported Deepfake of Maltese Prime Minister Robert Abela and Journalist Mark Laurence Zammit Used to Promote Fraudulent Investment

A deepfake scam allegedly used AI-generated audio to mimic Maltese Prime Minister Robert Abela and journalist Mark Laurence Zammit, falsely portraying them as endorsing an investment scheme. The fraudulent video repurposed genuine interview footage from 2022 but replaced the original dialogue with AI-generated voices.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.3 Fraud, scams, and targeted manipulation

The incident involves scammers using AI-generated voice clones of public figures to deceive individuals into signing up for a fraudulent investment scheme for financial gain.

Additional risk subdomains

• 3.1 False or misleading information: The deepfake generates highly realistic false statements and attributes them to public figures, misleading the audience.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The incident was caused by human scammers who intentionally created and deployed the audio deepfake to defraud social media users.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The incident involves AI-generated content, specifically an audio deepfake mimicking the voices of public figures, which falls under the category of deepfakes requiring transparency under the EU AI Act.

AI system and alleged parties

• AI system: None named
• AI purpose: Voice Generation; Deepfake Video Generation
• Behaviour type: Tool
• Alleged developer: Unknown voice cloning technology developers, Unknown deepfake technology developers
• Alleged deployer: Unknown scammers, Scammers impersonating Robert Abela and Mark Laurence Zammit
• Alleged harmed parties: Robert Abela, Mark Laurence Zammit, General public of Malta, General public, Epistemic integrity

Harm severity

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Minor, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Minor, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: Yes, the report explicitly describes a malicious audio deepfake created to impersonate public figures for a fraudulent scheme.

Directly caused: An audio deepfake mimicking the voices of Robert Abela and Mark Laurence Zammit was created and circulated on Facebook.

Indirectly caused: N/A

Inferred additional harm: The deepfake content likely reached thousands of Facebook users in Malta before being flagged or removed.

Epistemic

Reported: Yes, the report describes epistemic harm through the creation of a deepfake video that makes public figures appear to say things they never actually said.

Directly caused: A doctored video with fake English audio was superimposed over genuine interview footage to mislead viewers into believing the Prime Minister endorsed a trading platform.

Indirectly caused: N/A

Inferred additional harm: Such deepfakes contribute to the erosion of shared truth and make it harder for citizens to distinguish genuine media from manipulated content.

People affected

• Occurrences reported: 1
• People reportedly harmed: 2
• People reportedly exposed: 100

Potential causes

Management

• Poor Ad Safety Prioritization: Ad networks prioritize revenue over rigorous verification of ad content.
• Insufficient Risk Mitigation: Lack of proactive measures by platforms to prevent celebrity-based scams.

Technology

• Advanced Voice Cloning Tech: Scammers used AI to generate highly realistic voice deepfakes of targets.
• Video Manipulation Tools: Software allowed superimposing doctored audio onto genuine video footage.

Data Inputs

• Accessible Public Footage: Genuine interview video from social media was harvested to build the scam.
• Target Voice Samples: High-quality public audio allowed training of realistic voice models.

Human Factors

• Trust in Public Figures: Viewers trust the Prime Minister and journalists, making the scam effective.
• Limited Deepfake Awareness: Some users cannot easily distinguish manipulated audio from genuine clips.

Process and Methods

• Inadequate Ad Moderation: Facebook hosted the scam post without detecting the deepfake audio.
• Slow Platform Takedowns: Social media platforms let fraudulent ads run before taking action.

Regulatory Environment

• Weak Deepfake Regulation: Lack of strict legal penalties for creating and sharing deceptive deepfakes.
• Lack of Platform Accountability: No strong regulations forcing platforms to instantly verify ad authenticity.

Information quality

• Classification confidence: High
• Reason for confidence: The report from Times of Malta provides clear, direct evidence of the deepfake scam, identifying the source footage, the nature of the audio manipulation, and the fraudulent intent. There is little ambiguity about the event.
• Ambiguities identified: The specific AI tool or platform used to generate the deepfake is not identified.
• Alternative interpretations: None.

Scammers used AI audio deepfake technology to impersonate Maltese Prime Minister Robert Abela and a journalist in a fraudulent Facebook video promoting a fake investment scheme. The incident represents a minor national security concern, reflecting an established threat vector of financially motivated deepfakes targeting public figures, with negligible impact on core national security infrastructure or sovereignty.

• Overall national security impact: Minor
• Response level: Moderate
• Scope: Single nation
• Primary target: Malta
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. The immediate scam has been debunked by local media, representing an ongoing strategic concern regarding deepfake technology rather than an active national security crisis.
• Autonomy: Human-controlled. The AI voice-cloning tool was used as a tool by human scammers who directed the creation, content, and dissemination of the fraudulent video.
• Novelty: Established threat. Deepfake-enabled financial scams impersonating public figures and politicians are an established and well-documented global threat.

Impact by dimension

• Physical security: Negligible. No physical systems, critical infrastructure, or human safety elements were targeted or affected by this incident.
• Information security: Minor. While the incident involves a deepfake of a political leader, it is a financially motivated scam rather than a coordinated state-sponsored information warfare or intelligence compromise operation.
• Sovereignty: Minor. The Prime Minister's likeness was impersonated, which could subtly affect public trust, but there was no disruption to core government operations, decision-making, or electoral systems.
• Economic security: Minor. The scam targeted individual citizens for small financial deposits (250 euros) and does not pose a systemic threat to the nation's financial stability or strategic industries.
• Societal stability: Minor. The deepfake contributes to epistemic harm and potential loss of trust in media and public figures, but does not cause large-scale civil unrest or systemic human rights violations.