A new AI usage report has revealed that most enterprise AI security risks are connected to a small group of heavy AI users inside organizations. Researchers found that while many employees use AI tools occasionally, only a limited number of workers generate most of the risky AI activity. These users are often called “AI power users” because they spend much more time using AI systems. Experts say this group creates a much higher chance of sensitive business data being exposed.
The report explains that companies around the world are rapidly adopting generative AI tools for daily work activities. Employees now use AI for writing content, summarizing documents, coding assistance, research tasks, and data analysis. While these tools improve productivity and save time, organizations are still struggling to fully monitor employee AI usage. Researchers say many companies do not know exactly what information workers are uploading into AI platforms.
One major concern highlighted in the report is the accidental sharing of sensitive company information with AI systems. This may include internal business documents, customer information, source code, financial records, or intellectual property. Employees often upload this data without fully understanding the possible privacy or security risks connected to AI tools. Experts warn that organizations may lose control over sensitive data once it is shared with external AI platforms.
According to researchers, AI-related risks are not equally spread across all employees inside a company. Instead, a small number of “power users” are responsible for most high-risk AI activity within organizations. These employees spend significantly more time interacting with AI tools compared to normal users during daily work. Because of this increased usage, researchers believe they create a larger attack surface for possible data exposure incidents.
The report also discussed the growing problem of “shadow AI” in workplaces across different industries. This happens when employees use unauthorized AI tools or personal AI accounts instead of approved company systems. Researchers found that many workers prefer public AI tools because they are easier, faster, and more convenient to access. However, this creates major security concerns because IT teams may not know what data is being shared externally.
Security experts say enterprises are now facing new cybersecurity challenges as AI adoption continues growing very quickly. Although organizations are investing heavily in AI technologies, many companies are still developing proper governance and security controls. Experts believe innovation in AI is currently moving faster than risk management and cybersecurity protections inside businesses. This leaves many organizations exposed to privacy, compliance, and data protection risks connected to AI usage.
Researchers also found that most AI-related incidents involve data exposure rather than direct operational cyberattacks or system disruption. Personally identifiable information and intellectual property were identified as some of the most commonly exposed data types. Experts believe organizations must improve employee awareness and create stronger internal AI usage policies to reduce these risks. Better monitoring systems and controlled access to AI platforms were also recommended in the report.
Cybersecurity professionals say the report is an important reminder that enterprise AI security is not only about technology itself. Human behavior and employee activity patterns also play a major role in AI-related security and privacy risks. As businesses continue integrating AI tools into everyday operations, experts believe companies must focus more on responsible AI usage. Stronger visibility, governance, and internal security controls will become increasingly important in the future.
