Scam Waves, Deepfake Fraud and Refund Rules — 2026-09-13
This week saw a surge in high-value deepfake investment scams, including a £250,000 loss in the UK and a major indictment in Taiwan involving AI voice alteration. Regulatory responses intensified globally, with South Korea launching its first industry-wide prevention week and China leading a 46-nation coalition against cross-border fraud networks. Meanwhile, international law enforcement achieved significant breakthroughs with the extradition of romance scam suspects from South Africa to the US.
Scam Waves, Deepfake Fraud and Refund Rules — 2026-09-13
PSNI reports £250,000 loss to AI deepfake investment scam
On September 8, the Police Service of Northern Ireland (PSNI) reported that a victim lost £250,000 after falling for an investment scam featuring a fabricated video of a well-known financial figure. The fraudsters used the deepfake media to gain the victim's trust before taking remote access to their computer. This incident highlights the growing sophistication of AI-generated content in bypassing traditional visual security checks and underscores the urgent need for multi-factor authentication protocols in financial transactions.

Taiwan indicts 57 for NT$900m AI-powered romance scam
The Taipei District Prosecutors’ Office indicted 57 individuals, including a husband-and-wife team, for running a "pig butchering" romance scam that utilized voice-altering technology to defraud over 20,000 victims out of more than NT$900 million (US$28.34 million). The case, revealed in early September, demonstrates how scammers are integrating AI voice cloning into long-con schemes to maintain credibility over weeks or months. The scale of the indictment signals a shift toward prosecuting the technical infrastructure of these networks, not just the street-level operators.
Six Nigerian nationals extradited to US for $6M romance scam
In a significant cross-border enforcement action reported on September 11, six Nigerian nationals accused of swindling American women out of $6 million through online romance scams were extradited from South Africa to the United States. The suspects, linked to the "Black Axe" criminal network, face wire fraud and money laundering charges. This extradition marks a milestone in international cooperation against transnational fraud rings that have previously exploited jurisdictional gaps between Africa and North America.
Anthropic warns of AI-enhanced crypto scams targeting Indian investors
A September 2026 report by Anthropic highlighted that AI tools are making scams significantly more convincing, particularly for Indian crypto investors targeted via fake tax refund notifications and financial messages. The report notes that generative AI allows scammers to create highly personalized, localized lures at scale, increasing the success rate of phishing attacks. This development suggests that regulatory frameworks must evolve to address not just the scams themselves, but the accessibility of the AI tools enabling them.
Local view
South Korea launches first "Voice Phishing Prevention Week"
The Financial Supervisory Service (FSS) of South Korea initiated the country's first-ever "Voice Phishing Prevention Week" from September 7 to 13, involving ten major financial institutions. The campaign aims to raise awareness about evolving phishing tactics through branch visits, online content, and radio advertisements. While data from the Korean government indicates that voice phishing incidents dropped by 40% year-over-year due to previous crackdowns, local media like Asia Economy and E-Today emphasize that criminals are rapidly adapting to new technologies, necessitating this renewed industry-wide vigilance.
Japan’s National Police Agency releases July fraud statistics
On September 4, Japan’s National Police Agency released provisional statistics showing that special fraud damages reached ¥210.8 billion by the end of July 2026, a sharp increase of ¥63.2 billion compared to the same period last year. ScanNetSecurity highlighted that SNS-based investment fraud remains a primary driver of these losses. In response to the surge, local outlets like Saga Shimbun have issued specific warnings about SNS investment scams, urging citizens to verify official sources before engaging with any investment opportunities promoted on social media platforms.
Context & numbers
Global fraud losses and reimbursement trends
Recent data continues to frame the economic impact of these scams. In the UK, total fraud losses reached nearly £1.3 billion in 2025, with Authorized Push Payment (APP) fraud accounting for a significant portion. The Payment Systems Regulator (PSR) reported that 89% of eligible APP fraud victims were reimbursed under new rules implemented in late 2024, which contributed to a £73 million reduction in payment fraud losses. However, critics note that many victims remain out of pocket as rules are not consistently applied across all payment types.
FTC reports $3.5 billion in imposter scam losses
In the US, the Federal Trade Commission (FTC) reported that consumers lost $3.5 billion to imposter scams in 2025. The agency has intensified enforcement actions, including recent complaints against firms impersonating government agencies and insurance carriers. These figures underscore the persistent threat of social engineering, even as banks implement stricter verification protocols.
On the radar
- China leads 46-nation anti-fraud coalition: Reports from Asia Today indicate China is spearheading a new international alliance with 46 countries to combat cross-border voice phishing and online fraud centers. This geopolitical shift could lead to stricter information sharing and coordinated takedowns of scam hubs in Southeast Asia.
- Korean virtual asset refund rules take effect: Starting October 1, 2026, South Korea will implement new victim relief measures for virtual asset-related voice phishing, including freezing assets at the time of payment suspension. Investors should be aware of how these changes affect recovery prospects for crypto-based scams.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.
