Purportedly AI-Generated Deepfake Investment Ads Defrauded 5,000 Swedish Investors of 500 Million SEK

At least 5,000 Swedish investors were reportedly defrauded of around 500 million SEK (~$52.5 million USD) in 2025 through purportedly AI-generated deepfake investment ads circulated on Meta platforms. Fraudsters allegedly used fabricated videos and profiles of well-known finance figures such as journalist Gabriel Mellqvist and investor Günter Mårder to promote false stock tips in pump-and-dump schemes.

Risk classification

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

Fraudsters used AI-generated deepfakes of trusted financial figures to run pump-and-dump scams, defrauding over 5,000 investors of 500 million SEK.

Additional risk subdomains

• 6.3 Economic and cultural devaluation of human effort: The public likeness and professional reputation of well-known financial figures were exploited and devalued through unauthorized AI deepfakes.

Causal factors

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

The harm was caused by human fraudsters who intentionally deployed AI deepfake generation tools to deceive and defraud investors.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

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

AI system and alleged parties

• AI system: unspecified
• AI purpose: Deepfake Video Generation; Social Media Content Generation
• Behaviour type: Tool
• Alleged developer: Unknown voice cloning technology developers, Unknown deepfake technology developers
• Alleged deployer: Unknown scammers impersonating Günter Mårder, Unknown scammers impersonating Gabriel Mellqvist, Unknown scammers, Unknown pump-and-dump scammers
• Alleged harmed parties: Swedish investors, Günter Mårder, General public of Sweden, Gabriel Mellqvist, Epistemic integrity

Harm severity

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

Financial

Reported: The report explicitly describes total financial losses of 500 million SEK (half a billion kronor) affecting at least 5,000 investors.

Directly caused: N/A

Indirectly caused: Over 5,000 Swedish small savers lost a total of 500 million SEK in pump-and-dump schemes, averaging 100,000 SEK per victim.

Inferred additional harm: N/A

Malicious content

Reported: The report explicitly describes AI-generated deepfake videos and fake profiles used for scams.

Directly caused: Fraudsters created 15 to 20 AI-produced deepfake videos and fake profiles of financial experts to deceive users on social media platforms.

Indirectly caused: N/A

Inferred additional harm: N/A

Psychological

Reported: The report explicitly describes psychological distress and frustration experienced by the impersonated experts.

Directly caused: N/A

Indirectly caused: Financial expert [PERSON_008] described the experience as 'tough' and taking up his entire summer to handle, causing personal distress. Financial journalist [PERSON_001] expressed 'great frustration' over the identity theft.

Inferred additional harm: It is highly likely that many of the 5,000 defrauded investors experienced significant psychological distress, anxiety, and trauma due to losing large sums of money.

Epistemic

Reported: The report explicitly describes epistemic harm through the creation of highly realistic fake videos and profiles that undermine trust in digital platforms.

Directly caused: Fraudsters generated highly realistic deepfake videos of financial experts speaking in known environments to fabricate credible investment advice.

Indirectly caused: The incident contributed to an environment where it is 'difficult to trust anything within digital platforms,' as stated by [PERSON_008].

Inferred additional harm: N/A

People affected

• Occurrences reported: 1
• People reportedly harmed: 5002
• People reportedly exposed: 5002

Potential causes

Management

• Lack of Accountability by Meta: Meta management did not take responsibility or act against fake profiles.

Technology

• AI Deepfake Video Generation: AI was used to create realistic fake videos of trusted financial experts.
• AI Generated Fake Profiles: Fraudsters used AI to generate highly convincing fake profiles on social media.

Data Inputs

• Misuse of Public Media Footage: Scammers used existing videos of experts from known media settings to train AI.

Human Factors

• Overconfidence in Social Media Ads: Users trusted ads featuring well-known financial figures on Facebook.
• Lack of Digital Literacy in Elders: Older generations struggled to critically assess AI-generated video materials.

Process and Methods

• Inadequate Ad Moderation on Meta: Meta failed to detect and block deepfake scam ads on its platform.
• Use of Private WhatsApp Groups: Fraudsters moved victims to WhatsApp to coordinate pump and dump schemes.

Information quality

• Classification confidence: High
• Reason for confidence: The reports are consistent across multiple reputable Swedish media outlets, providing clear figures on financial losses and victims. The role of AI in generating the deepfakes is explicitly stated.
• Ambiguities identified: The specific AI tools or models used to generate the deepfakes are not identified.

A large-scale AI-enabled financial scam in Sweden defrauded 5,000 investors of 500 million SEK using highly realistic deepfakes of trusted financial experts. While not a direct threat to physical security or sovereignty, the incident represents a substantial economic security concern, showcasing how synthetic media can be scaled to cause significant financial harm and erode trust in digital systems.

• Overall national security impact: Substantial
• Response level: Substantial
• Scope: Single nation
• Primary target: Sweden
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents an ongoing strategic concern regarding the scaling of AI-enabled financial fraud rather than an active, imminent national security crisis.
• Autonomy: Human-controlled. The AI systems were utilized as tools by human fraudsters who made all operational decisions regarding target selection and financial execution.
• Novelty: Evolved capability. While deepfake fraud is an established threat, the scale, localized targeting, and coordinated execution mark a significant evolution in malicious capability.

Impact by dimension

• Physical security: Negligible. The incident did not involve kinetic attacks, physical threat, or manipulation of critical infrastructure.
• Information security: Minor. AI-generated deepfakes were used to impersonate trusted public figures, but the campaign was a financially motivated scam rather than state-sponsored information warfare.
• Sovereignty: Negligible. No core government operations, electoral systems, or sovereign decision-making processes were compromised.
• Economic security: Substantial. Significant financial impact with 5,000 Swedish retail investors losing 500 million SEK through coordinated AI-enabled market manipulation, representing a notable economic threat.
• Societal stability: Minor. The incident caused psychological distress to victims and impersonated experts, and eroded public trust in digital platforms, but did not cause widespread civil unrest.