AI is spreading faster than companies can secure it. A new cybersecurity report by CybSafe shows most users receive no training on the risks, even as sensitive data is routinely shared with chatbots and scams climb to record levels.
Training Gaps Widen as AI Use Accelerates
AI usage jumped 21% in the past year, with ChatGPT leading at 77% adoption, followed by Gemini at 49% and Microsoft Copilot at 26%, according to the National Cybersecurity Alliance and CybSafe. Yet 58% of users report no training on privacy or security implications, even as 43% admit to feeding AI systems internal documents, financial data, or client information without employer approval.
The findings point to a systemic blind spot. While AI is increasingly embedded into workflows, companies lag in establishing safe-use protocols. Cybercrime victimization is rising in parallel: 44% of respondents reported monetary or data loss from scams this year, a 9% jump from 2024. Younger groups, particularly Gen Z and Millennials, remain the most affected, with crypto fraud, phishing, and impersonation scams topping the list.
Basic Cyber Hygiene Still Falling Short
The report shows little improvement in fundamental security practices. Only 62% regularly create unique passwords, down from last year, and 41% never use a password manager. Multi-factor authentication, while widely recognized, is used consistently by fewer than half of respondents. Just 56% update software frequently and under half back up data regularly. Training availability has also stagnated: 55% say they have no access, and even when offered, usage is low.
Confidence in spotting malicious links remains uneven. Millennials are the most self-assured at 72%, while older generations trail significantly. Yet fewer than half of all participants report phishing attempts, undermining collective defense. At the same time, nearly two-thirds worry AI will make scams harder to detect and blur the line between authentic and fake information.
The Hidden Cost of Inaction
As AI tools become deeply integrated into business functions, unmanaged risk doesn’t only increase exposure to scams or data leaks, it raises the operational costs of remediation, compliance investigations, and trust rebuilding. Recent industry research shows the financial hit from a single data breach now averages more than $4 million globally. With AI accelerating both data generation and attack sophistication, the real price of inadequate training is likely to be measured in supply chain disruptions, higher insurance premiums, and lost commercial confidence.
