Managing AI in the Workplace: Why Browsers Have Become a Major Risk
In the modern technological landscape, AI governance has become a primary challenge for companies. While IT teams have successfully mastered the control of applications installed on Mac devices, the most significant link in data security remains vulnerable: the web browser.
Applications are controlled, but browsers are not
Today, most companies can accurately identify which AI programs are installed on their employees' computers. Using Mobile Device Management (MDM) systems, administrators can easily block or remove unwanted software. However, the problem arises when an employee does not install a specific program, but simply opens Safari or Chrome to access websites like ChatGPT or Claude.
When an employee logs into a personal account and shares corporate documents with an AI for analysis, the action does not appear on the list of applications monitored by the IT department. Monitoring data leakage at the web-application level is nearly impossible using traditional methods, leaving companies exposed to serious risks.
New approaches to security
Technology companies are seeking new ways to address this issue. For instance, Jamf plans to introduce AI management features at the browser level in the fourth quarter of 2026. This system will allow administrators to detect the use of personal AI accounts across various browsers, including Safari, Chrome, Firefox, and Edge, and block the sharing of sensitive information when necessary.
Another alternative is the use of so-called "enterprise browsers." For example, the company Island offers businesses a specialized browser that is fully managed by the IT department. In this environment, all actions occur within a container where administrators can define what types of data can be copied or uploaded when interacting with AI models. While this model is developing rapidly and attracting billions of dollars in investment, replacing the browser employees are accustomed to remains a difficult task, especially within the Apple ecosystem, where Safari is deeply integrated into the operating system.
What this means for businesses
For companies actively integrating AI into their workflows, this issue will become increasingly relevant in the near future. Data privacy protection should not be limited to a list of installed programs. When planning IT strategy, it is essential to review security policies at the web-browser level to prevent the accidental disclosure of corporate secrets through public AI models.
The security sector is evolving quickly, and the tools currently being developed may soon become the standard for organizations of any size.