Source URL: https://it.slashdot.org/story/25/04/18/209230/study-finds-50-of-workers-use-unapproved-ai-tools?utm_source=rss1.0mainlinkanon&utm_medium=feed
Source: Slashdot
Title: Study Finds 50% of Workers Use Unapproved AI Tools
Feedly Summary:
AI Summary and Description: Yes
Summary: The text discusses a study highlighting the prevalence of “Shadow AI” usage among employees, emphasizing the ease of access to AI tools and the challenges organizations face regarding security and compliance. It underscores the importance of understanding this phenomenon, as it presents various risks related to data exposure and usage of unsanctioned AI tools.
Detailed Description: The report by Software AG reveals crucial insights into employee behavior towards AI tools in the workplace, specifically focusing on the phenomenon of Shadow AI. Key points include:
– **Prevalence of Shadow AI**: About half of all employees are identified as users of Shadow AI, tools not officially sanctioned by their organizations. This trend indicates a significant shift towards informal usage of AI technologies.
– **Drivers of Use**: Employees are motivated to use AI tools for enhancing personal efficiency and career advancement, particularly in roles requiring quick access to information or content generation.
– **Ease of Access**: The report highlights how the ease of accessing AI tools contributes to rampant use, suggesting that when official tools are restrictive or difficult to access, employees turn to alternative sources, often using their personal accounts.
– **Low Malicious Intent**: The report posits that most employees do not have malicious intentions when using Shadow AI, but rather have a genuine desire to perform better in their roles.
– **Lack of Disclosure and Understanding**: The tendency for employees to not disclose the use of Shadow AI tools leads to organizations having limited visibility into the potential risks associated with these unsanctioned practices. The complex reasons behind this behavior include reluctance to appear less capable or fear of repercussions.
– **AI Usage Trends**: ChatGPT emerged as the most popular generative AI model used by employees, with a significant percentage of data prompts coming from personal accounts. Additional notable stats include:
– 68.3% of prompts involved image files.
– A small percentage of employees using Chinese AI models, indicating a diverse toolset.
– **Shifts in Data Sensitivity**: The report indicates notable changes in the types of sensitive information being exposed when using Shadow AI:
– Reduction in customer and employee data exposure.
– Increase in the exposure of legal, financial, and sensitive code data.
– Introduction of a new category tracking Personally Identifiable Information (PII) at a measured rate.
– **Implications for Security and Compliance**: The findings underscore the necessity for organizations to elevate their understanding and management of Shadow AI usage. This involves implementing robust security postures, educating employees about potential risks, and creating a more accessible supportive environment for needed AI tools.
In conclusion, the text reveals critical implications for security and compliance professionals, emphasizing the need to recognize Shadow AI as a significant risk factor and to formulate strategies that address this new landscape proactively.