Jerusalem, 10 December, 2025 (TPS-IL) — Tel Aviv–based AI security company DeepKeep unveiled a new high-precision guardrail designed to prevent personal data from leaking through enterprise AI systems. The tool detects, redacts and blocks sensitive information in prompts and outputs, expanding the company’s AI Firewall, which now includes more than 60 built-in guardrails.
DeepKeep says the capability offers broader and more accurate protection than existing tools and helps companies comply with global privacy regulations.
“Enterprises cannot afford to cut corners on AI security,” said CTO Yossi Atlevet. The feature is already available to customers.