Kosovo Public Broadcaster RTK and Journalist Gëzim Bimbashi Were Reportedly AI-Impersonated in Alleged Fake Joint-Cure Scam Video

Kosovo public broadcaster RTK's logo and journalist Gëzim Bimbashi's image and voice were reportedly AI-manipulated in a Facebook video promoting Versatil Pro as a supposed cure for joint disease.

Risk classification

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

The primary harm is the use of AI-generated deepfakes of public figures to scam consumers into buying unlicensed, overpriced medical products.

Additional risk subdomains

• 3.1 False or misleading information: The deepfakes spread false medical claims, such as curing chronic joint diseases or restoring sight in seven days.

Causal factors

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

The incident was driven by human actors intentionally deploying AI deepfake tools to commit financial fraud and deceive the public.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The incident involves AI-generated content such as deepfakes, which are subject to transparency obligations under the Act.

AI system and alleged parties

• AI system: unspecified
• AI purpose: Deepfake Video Generation; Voice Generation
• Behaviour type: Tool
• Alleged developer: Voice cloning technology developers, Synthetic audio technology developers, Deepfake technology developers
• Alleged deployer: Scammers in Kosovo, Scammers, Deepfake creators
• Alleged harmed parties: Social media users in Kosovo, Social media users, Radio Television of Kosovo, People seeking medical advice, Journalistic integrity, Gëzim Bimbashi, General public of Kosovo, General public, Facebook users in Kosovo, Facebook users, Epistemic integrity

Harm severity

Severity ratings are based on this scale.

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 Minor, indirect Negligible
• Epistemic: direct Minor, indirect Substantial
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: The report explicitly describes the creation and spread of malicious deepfake videos and fabricated articles.

Directly caused: At least 83 AI-generated deepfake videos and articles were created and spread across social media platforms like Facebook.

Inferred additional harm: Widespread exposure of the general public in Kosovo and Albania to deceptive, malicious advertising content on social media.

Psychological

Reported: The report explicitly describes distress and reputational concern among the targeted public figures.

Directly caused: At least 7 named public figures experienced distress and reputational risk due to their likenesses being stolen and associated with fraudulent medical products.

Inferred additional harm: Scammed consumers likely experienced distress and anxiety upon realizing they purchased fake, potentially dangerous medical products.

Epistemic

Reported: The report explicitly describes extensive epistemic harm through the fabrication of medical claims and fake news broadcasts.

Directly caused: 83 AI-generated articles and videos fabricated medical breakthroughs, fake interviews, and fake WebMD-style reports, severely misleading the public.

Indirectly caused: The misuse of trusted media logos (RTK, KTV, Top Channel, TV Klan) and professional identities erodes public trust in legitimate journalism and healthcare institutions.

Inferred additional harm: A broader degradation of the shared information ecosystem regarding public health and medical science in the region.

People affected

• Occurrences reported: 83
• People reportedly harmed: 7
• People reportedly exposed: 10000

Potential causes

Technology

• AI Voice and Video Cloning: AI cloned voices and manipulated video of news presenters to deceive viewers.
• Virtual Machine Evasion: Scammers used virtual machines to bypass Meta location and monetization blocks.

Data Inputs

• Misused Broadcast Footage: Authentic news footage was used as input to generate realistic deepfakes.
• Stolen Public Identities: Names and images of real doctors and artists were fed into AI scam engines.

Human Factors

• Public Trust in TV Logos: Viewers trusted deepfakes because they displayed familiar television logos.
• Low Digital Literacy: Audiences failed to recognize AI-generated speech and video manipulations.

Process and Methods

• Weak Social Media Moderation: Platforms failed to proactively detect and remove AI-generated scam videos.
• Inefficient Fact Checking: Fact-checking occurred after scams spread, failing to prevent initial harm.

Regulatory Environment

• Regulatory Authority Gaps: Medical agencies lacked the power to ban supplement ads on social media.
• Outdated Supplement Laws: Lack of specific laws for food supplements allowed online sales loopholes.
• Lagging Regional Alignment: Slow progress in aligning cross-border healthcare and medicine legislation.

Information quality

• Classification confidence: High
• Reason for confidence: The reports from Krypometer provide clear, detailed, and verified facts regarding the deepfake campaign, including specific numbers of AI-generated articles (83 out of 220) and direct statements from the affected individuals.
• Ambiguities identified: The specific AI models or software used to generate the deepfakes are not identified.
• Alternative interpretations: None. The evidence clearly points to a coordinated commercial fraud campaign using synthetic media.

A widespread commercial fraud campaign in Kosovo and Albania utilized AI-generated deepfakes of doctors and journalists to sell unlicensed medical products. While eroding epistemic trust in media and posing minor public health risks, the incident represents a low-level threat with minor national security implications.

• Overall national security impact: Minor
• Scope: Multiple nations
• Primary target: Kosovo and Albania
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. The campaign represents an ongoing strategic concern regarding regional disinformation and fraud rather than an active crisis requiring immediate state response.
• Autonomy: Human-controlled. The AI systems were used as tools by human operators to generate synthetic media, with humans controlling the distribution and monetization processes.
• Novelty: Evolved capability. While deepfake scams are established, the highly localized targeting of specific regional public figures and media brands represents an evolved threat vector.

Impact by dimension

• Physical security: Negligible. No kinetic attacks, critical infrastructure disruption, or direct physical casualties were reported in connection with this incident.
• Information security: Minor. AI-generated deepfakes impersonated journalists and fabricated media logos to spread disinformation. However, this was for commercial fraud rather than state-sponsored information warfare.
• Sovereignty: Negligible. The incident did not compromise government decision-making, electoral systems, or state sovereignty.
• Economic security: Minor. Scammers bypassed platform monetization restrictions and sold unlicensed products at inflated prices, causing minor localized economic harm but no strategic industry threat.
• Societal stability: Minor. The exploitation of trusted public figures and media outlets erodes public epistemic trust, though it does not represent mass surveillance or systemic rights violations.