AI Use at Work Is Creating Security Risks

A new report from Deloitte shows employees often use public AI tools without their employer's approval, increasing the risk of trade secrets leaking. Companies should adopt guidelines rather than bans, says AI expert Patric Barenthin.

Share
AI Use at Work Is Creating Security Risks

A new study from Deloitte shows employees are using AI at work without guidelines or policies from their employers. The majority, a full 52 percent, use publicly available free tools, while 25 percent pay for their own AI services. Only 19 percent use AI tools provided by their employer or a third-party vendor.

Risk of Trade Secrets Leaking

Patric Barenthin, head of AI and data analytics at Deloitte Sweden, warns that employees' use of public AI tools poses a major security risk. He points out that public AI models are trained on user data, which could lead to sensitive information being exposed.

"Employees don't intend to leak trade secrets, but if they use public AI services, the information can become accessible to others. It's hard to remove data from these services once it's been uploaded," says Barenthin.

Guidelines Are the Solution, Not Bans

Instead of banning AI use, Barenthin recommends companies develop clear guidelines. He argues it's the employer's responsibility to educate employees about the risks, and that employees, in turn, should avoid entering sensitive information into public AI tools.

"Companies should give their employees simple, clear guidelines to reduce the risk of unintentional information leaks," he says.