Sensitive Enterprise Data Uploads to AI Models Increased 93% in a Year, Zscaler Report Finds (infosecurity-magazine.com)

• 93% increase in sensitive data uploads to AI tools year-over-year
• 18,033 TB of data transferred to AI/ML applications
• Grammarly (38%) and ChatGPT (21%) drove over half of transfers
• 410 million ChatGPT DLP violations, up 99%
• 242 million Codium DLP violations, up 100%

"A Zscaler report found a 93% year-over-year increase in employees uploading sensitive enterprise data to AI tools, with over 18,000 TB transferred. Grammarly and ChatGPT accounted for over half of transfers. DLP violations related to ChatGPT exceeded 410 million, a 99% increase."