United States: Even a small reduction in sleep could make people more vulnerable to cyber fraud, with new research finding that employees were significantly more likely to click on suspicious links after losing just one hour of sleep.
Researchers at the University of California, Irvine, studied the impact of sleep deprivation on cybersecurity awareness by examining the transition to Daylight Saving Time in March 2026. The findings, published in the Proceedings of the National Academy of Sciences (PNAS), suggest that even temporary sleep disruption can affect judgement and decision-making.
The study analysed responses from more than 11,000 employees at a large US institution who regularly receive simulated phishing emails as part of a cybersecurity training programme. These messages are designed to replicate fraudulent emails and assess whether recipients click on potentially harmful links.
Researchers found that on the Monday following the clock change, 17.76% of employees clicked on the simulated phishing links, compared with 15.56% on the preceding Monday. This represented a 13% increase in the likelihood of falling for the simulated attacks.
Lead researcher Dr Eitan Schechtman, assistant professor of neurobiology and behaviour at UC Irvine, said the findings demonstrate how even limited sleep loss can influence everyday decisions, including those involving cybersecurity.
The research also raises concerns for businesses and individuals as cybercrime continues to cause substantial financial damage. According to the FBI’s Internet Crime Complaint Center, reported cybercrime losses in the United States exceeded $20 billion in 2025.
Researchers said the findings could help employers strengthen cybersecurity precautions during periods when sleep patterns are disrupted, including clock changes, holidays and major sporting events.
The study also adds to the debate over permanent clock arrangements in the US, with researchers favouring Standard Time because it better corresponds with the body’s natural biological rhythm.
