- Three in five workers say they use unprecedented AI tools (shadow)
- Managers and senior executives are the worst culprits
- Workers do not have the right tools or policies
New research has said that three in five workers (59%) say they are using AI tools that have not been approved by their business – otherwise known as Shadow AI.
Worse still, 75% of people who use a shadow AI admit to share sensitive data that could endanger their businesses, and 57% of direct managers of these employees support the use of an unprecedented AI, figures for figures for figures Cyberness say.
Indeed, it is in fact executives and senior managers who are most likely to use IA Shadow tools (93%), with managers (73%) and professionals (62%) less likely to do so.
Shadow AI is a great concern for organizations
Among the most commonly shared sensitive information is employee data (35%), customer data (32%), internal documents (27%), legal and financial information (21%), security -related information (21%) and the proprietary code (20%), despite most (89%) workers associating AI with risks.
Almost two -thirds (64%) recognize that data violations could result from the use of shadow AI, but while 57% are suitable, they would stop using unprecedented tools if a data violation occurred, few take preventive measures today.
“Once the sensitive data is part of an unmarked AI tool, you lose control. It can be stored, reused or exposed in a way you will never know,” said product manager Nexos.ai ž žilvinas Girėnas.
Although companies are trying to control the use of the shadow AI, a quarter (23%) still has no official AI policy. Similarly, only half (52%) of employers offer AI tools approved for work, and only one in three workers consider them to meet their needs.
Obviously, it is up to companies to implement more robust policies and to offer the right type of tool that meets the needs of their workers – not only generic.
“Companies should study the means to integrate AI into their safe, efficient and responsible processes”, ” Cyberness Security researcher Mantas Sabeckis concluded.
