“People are embracing AI in their personal and professional lives faster than they are being educated on its risks,” said Lisa Plaggemier, Executive Director of the NCA. The study revealed that 43% of respondents admitted to sharing sensitive information, including company financial data and client records, with AI chatbots, often without realizing the potential consequences. The findings highlight a growing disconnect between AI adoption and cybersecurity preparedness, suggesting that many organizations are failing to educate employees on how to use these tools responsibly.
The NCA-CybNet report aligns with previous warnings about the risks posed by AI systems. A survey by software company SailPoint earlier this year found that 96% of IT professionals believe AI agents pose a security risk, while 84% said their organizations had already begun deploying the technology. These AI agents—designed to automate tasks and improve efficiency—often require access to sensitive internal documents, databases, or systems, creating new vulnerabilities. When improperly secured, they can serve as entry points for hackers or even cause catastrophic internal errors, such as one case where an AI agent accidentally deleted an entire company database.
Traditional chatbots also come with risks, particularly around data privacy. Despite assurances from companies, most chatbot interactions are stored and sometimes used for future model training, meaning they are not entirely private. This issue gained attention in 2023 when Samsung engineers accidentally leaked confidential data to ChatGPT, prompting the company to ban employee use of the chatbot.
The integration of AI tools into mainstream software has only accelerated their ubiquity. Microsoft recently announced that AI agents will be embedded into Word, Excel, and PowerPoint, meaning millions of users may interact with AI daily—often without any specialized training in cybersecurity. As AI becomes an integral part of workplace tools, the potential for human error, unintentional data sharing, and exposure to security breaches increases.
While the promise of AI continues to drive innovation, experts warn that its unchecked expansion poses significant security challenges. Without comprehensive training, clear policies, and safeguards in place, individuals and organizations risk turning powerful productivity tools into major sources of vulnerability. The race to integrate AI into every aspect of modern life is well underway—but for cybersecurity experts, the race to keep users informed and protected is still lagging far behind.
