Deepfakes shatter online confidence

TechnologyDigital
10 May 2026 • 12:00 AM MYT
The Manila Times
The Manila Times

One of the longest-running English broadsheets in the Philippines

Deepfakes shatter online confidence

A NEW consumer study from iProov, a leading provider of science-based biometric identity verification solutions, warns that the rise of deepfakes is undermining confidence in what people see online. As a result, artificial intelligence (AI)-driven impersonation is increasingly being perceived as a real-world threat.

Nearly half (48 percent) of respondents in the study now question the authenticity of “almost everything” they encounter online, indicating a profound shift in consumer trust. Consequently, genuine human presence in digital ecosystems is now essential to thwart fraud, misinformation, reputational harm and institutional distrust, and to regain certainty about what — or who — can be believed.

The breakdown in confidence is already spiraling into commercial pressure: Three-quarters (74 percent) of consumers say they would switch banks if a competitor offered stronger safeguards against deepfake-enabled fraud. The serious impact on consumer allegiance is clear, with younger adults — the most digitally native generation — leading this shift.

At the same time, expectations of accountability are rising sharply. More than half of respondents believe banks should be legally liable for losses caused by AI-enabled deepfakes, reflecting mounting demand for institutional responsibility in the AI era.

The study also highlights growing expectations around online government platforms. Many consumers said they would be more likely to use government services online if a secure biometric platform were available. Furthermore, respondents said they would prefer a secure face scan through a mobile app over an in-person appointment.

The stakes behind this research are easier to understand when experienced firsthand. iProov's new interactive game, “Find the Fake,” challenges participants to identify a single AI-generated deepfake hidden among a grid of six seemingly real social media profiles. It is a deceptively simple yet revealing exercise. Readers may try it at https://quiz.iproov.com/quiz2/en

The study surveyed 2,000 individuals across the United Kingdom and the United States in the first quarter of 2026.

View Original Article