AI-Generated Fake Content Drives Surge in Online Fraud in Italy
A Visa study found that Italian consumers unable to identify AI-generated fake content are five times more vulnerable to online fraud. The phenomenon has reportedly driven an estimated 7.4 million Italians to reduce or stop online shopping, with an estimated economic cost of €359.4 million annually.
What happened
According to a Visa study, consumers in Italy who cannot distinguish AI-generated fake content from real content are five times more likely to fall victim to online fraud, with a 59% vulnerability rate compared to 12% for those who can identify fakes.
The study found that 50% of Italian social media users noticed an increase in suspicious or potentially fraudulent advertisements in the past year, with deepfake videos and AI-generated synthetic content being used as fraud tools.
Reported impact
The study estimates that 7.4 million Italians have reduced or completely stopped making online purchases after experiencing or being targeted by fraud attempts, with 47% of fraud victims suffering direct economic damage averaging €147 per incident.
The phenomenon has created an estimated annual economic cost of €359.4 million for Italy, and 47% of affected users spent more than 24 hours resolving fraud-related issues after attacks.
Index analysis
The Visa study reveals that 28% of Italians believe artificial intelligence will make fraud increasingly difficult to detect on social media, while 12% of respondents admit to not adopting any online protection measures.
Visa launched an educational campaign using AI-generated social media posts to test user readiness, with results showing many consumers interacted with obviously implausible fake content.
Governance controls
Visa stated it is implementing AI-based tools to identify and block fraudulent transactions before they reach consumers, in collaboration with banks, retailers and platforms.
Visa collaborated with Olympic freestyle skier Leonardo Donaggio for an awareness campaign to help consumers recognize AI-generated fraud and adopt protective measures.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.