Sydney, Australia – Employee trust in digital communications is increasing despite growing risks from AI-generated impersonation, with more than half of workers failing to independently verify senders before acting on sensitive requests, according to research commissioned by WatchGuard Technologies.
The global survey of more than 500 full-time employees across eight countries found that 46% trust digital communications more than they did five years ago, while 31% said their trust had declined. However, 55% said they do not always independently verify the sender before responding to requests involving payments, credentials or confidential information.
The findings highlight how familiar names, voices, faces and branding can influence employees’ decisions, even as AI tools make these signals easier to imitate.
“The most striking finding isn’t that employees trust digital communications. It’s that many continue to trust the same signals they’ve relied on for years, even as AI makes those signals easier to imitate,” said Corey Nachreiner, Chief Information Security Officer at Watchguard.
“Familiar names, voices, faces, and brands have always helped people make quick decisions, but they’re no longer reliable indicators of authenticity. Verification is becoming one of the most important cybersecurity skills employees can have,” Nachreiner continued.
The research found that 72% of respondents had clicked a link or opened an attachment because it appeared to come from someone they knew. A known sender was cited by 38% as a reason to trust a digital request, while 36% pointed to familiar workplace or personal details and 32% to correct company branding. Only 27% said independent verification influenced their trust in a request.
The findings also highlighted potential vulnerabilities involving urgent requests from senior executives. Some 42% of respondents said they would likely trust an urgent text from their chief executive or manager requesting payment or sensitive information without verification. A similar voice message would be trusted by 40%.
Among respondents in the United States, 59% said they would trust such a text message.
When assessing unusual or sensitive video requests, 53% said they rely on a person’s face, voice or display name to verify their identity. By comparison, 28% prioritise checking through another communication channel or following an established organisational procedure.
38% of respondents identified a lack of employee training and awareness as their organisation’s biggest obstacle to cybersecurity preparedness.
Verification habits also varied across countries and industries. Australia recorded the highest verification rate among the countries surveyed, at 69%, while 22% of respondents in Germany said they check the sender. Respondents in Italy identified a shortage of cybersecurity skills and qualified staff as their leading obstacle, while those in Spain pointed to gaps in organisational security training.
By industry, healthcare employees reported the highest verification rate at 63%, while financial services respondents recorded the highest level of trust in digital content. Government employees were among the most sceptical of digital communications and executive text messages, although fewer than half said they verified sensitive requests.
The findings point to a role for managed service providers (MSPs) and channel partners in helping organisations improve verification practices and strengthen AI governance. Recommended measures include establishing confirmation procedures for payments, credential resets and confidential data sharing, as well as conducting realistic executive impersonation exercises.
Organisations are also encouraged to define approved AI tools, set data-handling rules and establish requirements for human review. The research further highlights the importance of accessible incident-reporting channels and tracking employee verification behaviour alongside technical security performance.
