{
  "id": 14384257,
  "title": "Over a third of Singapore workers unsure they can spot CEO deepfakes, says KnowBe4 study",
  "url": "https://urgent.news/2026/10/14/over-a-third-of-singapore-workers-unsure-they-can-spot-ceo-deepfakes",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-14T06:33:13.000Z",
  "source": {
    "name": "Malay Mail",
    "slug": "malay-mail-malaymail",
    "url": "https://www.malaymail.com/news/singapore/2026/10/14/over-a-third-of-singapore-workers-unsure-they-can-spot-ceo-deepfakes-says-knowbe4-study/239070"
  },
  "original_language": "en",
  "account": "A KnowBe4 study has revealed that over a third (36 per cent) of Singapore office workers lack confidence in distinguishing authentic messages from their chief executive officer (CEO) from those created through artificial intelligence (AI), including voice cloning and deepfake technology. This finding is detailed in the report \"Would You Know Your CEO’s Real Voice? Singapore's Growing Deepfake Vulnerability,\" which is based on a Censuswide survey of 500 office workers in Singapore. The report was unveiled during Singapore International Cyber Week and GovWare.\n\nDr Kawin Boonyapreede, Chief Information Security Officer advisor at KnowBe4, emphasized that Singapore's hierarchical structure creates a \"trust shortcut\" that AI-driven scams are exploiting. He noted that a cloned voice or a seemingly genuine message from \"the boss\" does not need to deceive technology; it merely needs to trigger an instinct not to question authority. To combat this, Boonyapreede suggested that organizations should provide training and simulations to alter behavior across all levels of seniority.\n\nThe research uncovered a dual vulnerability within Singapore workplaces. Junior and middle-tier employees often hesitate to question senior leaders, while senior managers may be overly confident in their ability to spot deepfakes. Specifically, 50 per cent of respondents expressed discomfort in challenging a request from a senior executive, whereas 52 per cent were concerned about potential repercussions for doing so. Additionally, 45 per cent were equally apprehensive about delaying a senior leader's request as they were about inadvertently causing a cybersecurity incident, with 9 per cent even preferring to risk a security breach over delaying their boss's request.\n\nSenior managers were the most confident in their ability to identify fake messages, with 78 per cent expressing such confidence. However, they were also the most likely to comply with a cloned CEO voice before verifying it, at 35 per cent, compared to 16 per cent among clerical workers. Only 26 per cent of respondents had undergone training focused on AI-generated scams involving senior leaders. Moreover, only 10 per cent identified the over-reliance on senior leaders as the greatest cybersecurity risk in their workplace, indicating that this vulnerability might be under-recognized.\n\nOrganizations, according to the report, need more than just technology to address the growing threat of AI-enabled impersonation targeting senior leaders. Boonyapreede stressed the importance of fostering an environment where junior and middle-tier employees feel safe to raise concerns and encouraging senior managers to implement stronger verification practices instead of solely depending on intuition.",
  "summary": "KUALA LUMPUR, Oct 14 — KnowBe4, a digital workforce security company, has found that more than a third (36 per cen...",
  "key_points": [
    "36% of Singapore office workers unsure about CEO authenticity",
    "Hierarchical trust exploited by AI-driven deepfakes",
    "Training needed to combat senior leader impersonation risk"
  ],
  "editors_take": "Singapore's hierarchical workplace culture creates a dangerous trust shortcut that allows AI-driven impersonation scams to exploit employees' reluctance to question authority or delay requests from senior leaders.",
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}