DeepKeep has introduced a new feature to protect personal information, called a PII guardrail, which scans as well as filters personal data in both prompts and responses across AI systems used in businesses.
This guardrail is part of DeepKeep's AI Firewall, which already includes over 60 built-in security controls. The company is focusing on large organizations that use generative AI and machine learning tools for various business tasks.
The PII guardrail works by detecting, removing, and blocking personal information in AI workflows. It looks at both the prompts (requests) from employees and the answers generated by AI models. This feature aims to reduce the risk of accidental data leaks as AI is used more widely in businesses.
As AI becomes more common, companies have been worried about sensitive data being accidentally shared. Employees often put internal information into AI tools, and AI can sometimes "guess" or reveal personal data based on what it has learned from past data.
Executive Statement
According to Yossi Atlevet, Chief Technology Officer at DeepKeep, as reliance on AI continues to grow exponentially, enterprises are increasingly reminded that they cannot afford to cut corners on AI security. Their PII guardrail allows businesses to reclaim their power over data security, with a faster and more detailed layer of protection, ensuring that privacy and compliance are built into every stage of AI usage. With a continuous, intelligent layer of defense, this guardrail affords enterprises the ability to innovate responsibly and protect their employees' and their customers' data and trust.