Avast Report Highlights AI Confidence Trap in Data Sharing and Deepfake Detection

• •56% of adults are confident in spotting AI deepfakes.
• •52% have shared sensitive data with AI tools, indicating a significant risk.
• •The report warns of an 'AI confidence trap' affecting data security.

The Avast 2026 Safe Tech Report reveals a significant gap in public perception regarding AI deepfakes and data sharing. It found that 56% of adults globally believe they can easily identify AI-generated fakes, while 52% have shared sensitive information with AI chatbots. This overconfidence creates an 'AI confidence trap,' where individuals feel secure in their ability to detect scams but are unaware of the risks associated with sharing personal data. The report indicates that AI was involved in millions of video encounters in early 2026, with about 1 in 30 flagged as scams. The findings suggest that while technology evolves, the psychological tactics used by scammers remain consistent. The report emphasizes the need for individuals to be cautious about sharing personal information with AI tools, as these interactions can lead to long-term data exposure. The study was commissioned by Avast to support its mission of promoting digital safety.