Alleged Deepfake Investment Scam Uses Economist David Rosenberg's Likeness on Meta Platforms

Fraudsters allegedly used AI-generated content and social media ads to impersonate economist David Rosenberg and promote a fraudulent investment scheme. Victims were reportedly directed to WhatsApp groups with fake stock tips, resulting in substantial financial losses, with some exceeding $500,000. The ads reportedly appeared on Meta platforms, including Facebook and Instagram. Rosenberg and his family reported the scam to authorities.

Risk classification

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

The incident involves malicious actors using AI-generated deepfakes of financial experts to conduct targeted investment scams and fraud, resulting in significant financial losses for victims.

Causal factors

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

The incident was caused by human fraudsters intentionally deploying and utilizing AI deepfake generation tools to scam investors.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Risk Level 3: Limited Risk. The reports describe the use of AI-generated content such as deepfakes to impersonate financial experts, which falls under transparency obligations for synthetic media.

AI system and alleged parties

• AI system: unspecified
• AI purpose: Deepfake Video Generation; Voice Generation
• Behaviour type: Tool
• Alleged developer: Unknown voice cloning technology developers, Unknown deepfake technology developers
• Alleged deployer: Unknown scammers impersonating David Rosenberg, Unknown scammers
• Alleged harmed parties: Unnamed woman who lost $500,000, Unnamed man who lost $450,000, Syed Hasan, Investors, General public, David Rosenberg

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Substantial. 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 Substantial, 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

Financial

Reported: The report explicitly describes total losses exceeding $1 million for one firm's contacts, with specific individual losses of $16,000, $500,000, and $450,000. Average financial loss per occurrence across the three detailed campaigns is estimated at $333,333.

Directly caused: Victims lost money directly to the fraudulent schemes promoted by the AI deepfakes, including individual losses of $16,000, $500,000, and $450,000.

Inferred additional harm: Given the rising trend of social media investment scams in Canada reaching $128.4 million, actual financial losses from these deepfake campaigns are likely much higher than the reported $1 million.

Malicious content

Reported: The report explicitly describes malicious deepfake content created and spread directly by the incident.

Directly caused: Fraudsters created hyper-realistic deepfake videos and cloned voices of financial experts to deceive users on Facebook and Instagram.

Inferred additional harm: It is likely that many more deepfake ads were generated and distributed across various social media platforms, reaching millions of users.

Epistemic

Reported: The report explicitly describes epistemic harm caused directly by the incident through the creation of hyper-realistic deepfakes of real financial experts giving fake advice.

Directly caused: The deepfakes fabricated statements and investment advice from real economists, misleading viewers into believing the experts endorsed these schemes.

Inferred additional harm: The proliferation of such deepfakes erodes public trust in legitimate financial commentators and digital media platforms.

People affected

• Occurrences reported: 3
• People reportedly harmed: 10
• People reportedly exposed: 1000000

Potential causes

Management

• Delayed Deployment of Safeguards: Meta only deployed facial recognition verification after major public pressure.
• Prioritizing Ad Revenue Over Safety: Platforms allowed unverified ads to run without rigorous pre-screening.
• Inadequate Risk Assessment of AI: Management failed to anticipate the scale of AI-driven imposter fraud.

Technology

• AI Deepfake Video Generation: Hyper-realistic AI videos realistically mimicked financial experts' movements.
• AI Voice Synthesis: AI synthesized realistic voices to make fraudulent investment advice convincing.
• Internet Scraping Tools: Scammers scraped legitimate business data to train and build fake profiles.

Data Inputs

• Lack of Ad Identity Verification: Meta did not verify the identity of advertisers using public figures' names.
• Scraped Public Media Data: Abundant public video and audio of experts was used to train deepfake models.
• Unlabeled AI Generated Content: AI-generated ads lacked metadata or labels indicating they were synthetic.

Human Factors

• Investor Trust in Authority Figures: Victims trusted the realistic AI likeness of well-known financial experts.
• Desire for High Yield Returns: Low interest rates drove investors to trust unrealistic high-yield promises.
• Inability to Detect Deepfakes: Users could not distinguish highly realistic AI avatars from real videos.

Process and Methods

• Slow Ad Takedown Procedures: Meta's manual and slow review processes allowed scam ads to run for months.
• Ineffective Ad Moderation Algorithms: Automated moderation failed to flag obvious deepfake scams targeting users.
• Whack-a-Mole Reporting System: Reporting scams was reactive and failed to prevent re-uploading of fake ads.

Regulatory Environment

• Safe Harbour Legal Protections: Platforms face limited liability for user-generated scam ads under safe harbor.
• Lack of National Fraud Agency: Canada lacks a unified, proactive fraud-fighting agency like Australia.
• Cross-Jurisdictional Police Barriers: International scammers operate outside the reach of local law enforcement.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide clear, consistent, and detailed accounts of multiple deepfake impersonation scams targeting financial experts, including specific financial losses and platform reach. The role of AI in generating these videos and voices is explicitly stated.
• Ambiguities identified: The specific AI models or software used to generate the deepfakes are not identified.
• Alternative interpretations: None. The events clearly represent a coordinated financial scam utilizing synthetic media.

Fraudsters used AI-generated deepfakes and voice cloning to impersonate prominent financial experts, defrauding Canadian and international investors of over $1 million. While representing an evolved cybercrime capability with significant individual financial harm, the incident remains a criminal fraud issue with minor national security implications.

• Overall national security impact: Minor
• Scope: Multiple nations
• Primary target: Canada
• Other affected: European Union
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents an ongoing strategic concern regarding the proliferation of synthetic media in financial fraud rather than an active national security crisis.
• Autonomy: Human-controlled. AI tools were directly controlled by human fraudsters to generate synthetic media, with humans managing the subsequent WhatsApp scam groups.
• Novelty: Evolved capability. Represents an evolved capability where high-quality generative AI and voice cloning are increasingly accessible and deployed at scale for targeted financial fraud.

Impact by dimension

• Physical security: Negligible. No physical systems, infrastructure, or human safety threats were reported in connection with this financial fraud incident.
• Information security: Minor. AI-generated deepfakes of financial experts were distributed to nearly one million users, representing a minor information security issue but lacking state-sponsored information warfare characteristics.
• Sovereignty: Negligible. The incident involves commercial financial fraud and does not impact government decision-making, elections, or sovereignty.
• Economic security: Minor. While individual victims suffered substantial financial losses totaling over $1 million, the incident does not threaten national economic stability, critical financial infrastructure, or strategic industries.
• Societal stability: Minor. The widespread use of hyper-realistic deepfakes to defraud citizens risks eroding public trust in digital media and financial communications, but does not constitute systematic oppression or mass civil unrest.