The AI Visibility Gap
The report from LayerX Security reveals a significant disparity in how artificial intelligence is used across enterprises. While nearly half of all employees have interacted with AI tools in the past year, only 18 percent use them on a weekly basis. This means most employees remain casual users, which might initially seem reassuring to security teams. However, the data shows that risk is not evenly distributed. A small fraction of users, identified as AI power users, generate a disproportionately high volume of activity.
Concentrated Activity and Deep Engagement
The top 5 percent of users generated at least 144 conversations each, while half of all users had 12 conversations or fewer. Power users also engage in much deeper interactions, averaging 18 prompts per conversation compared to the average of 2. This creates a new risk profile where a relatively small group of employees drives the majority of sensitive data exposure. The report underscores that most organizations still lack full visibility into this fragmented AI ecosystem, which includes personal accounts, browser extensions, embedded copilots, and tools operating outside traditional governance controls. ChatGPT remains the dominant platform, but the rapid fragmentation across multiple tools amplifies the challenge.
Source: The Hacker News
