AI has become a critical vector for corporate data leakage, surpassing shadow SaaS and unmanaged sharing channels. Immediate and comprehensive governance strategies are needed as sensitive data already flows into AI platforms through unmonitored accounts and copy/paste activities. #LayerX #AIDataLeakage
Keypoints
- AI tools are now widely adopted in enterprises, with 45% of employees using generative AI apps.
- Most AI sessions occur outside of enterprise control, mainly through unmanaged personal accounts.
- 40% of files uploaded to GenAI contain sensitive information like PII or PCI data.
- Copy-paste activity into AI platforms is the primary channel for data leakage, with 82% from unmanaged accounts.
- Security strategies must shift from file-based to action-based, focusing on AI, chat, and high-risk SaaS categories.
Read More: https://thehackernews.com/2025/10/new-research-ai-is-already-1-data.html