NEARLY 10% OF EMPLOYEE GEN AI PROMPTS INCLUDE SENSITIVE DATA

Summary from the TLDR InfoSec newsletter

Harmonic, an AI data security company, revealed that 8.5% of employee prompts to popular LLMs included sensitive data. Harmonic's report showed that 46% of submitted sensitive data included billing and authentication data, 27% included employee PII and payroll data, and 15% included legal and financial data. Employees who enter sensitive data to free tiers of LLMs may be inadvertently leaking data, as the report demonstrated that 54% of sensitive prompts were entered on ChatGPT's free tier.