DeepKeep

AI security product combining a firewall, usage visibility, and trust controls for LLM and vision models.

DeepKeep protects AI applications across their lifecycle, screening inputs and outputs with a firewall, monitoring how models are used, and evaluating them for weaknesses. It covers both language and computer vision systems. The Israeli company was founded in 2021.

Category: AI Model Security & Red Teaming
Founded: 2021
Funding: $24M raised (Grant, 2025)
Website: https://www.deepkeep.ai

Seen at Black Hat USA 2026 (Startup City sponsor, booth 5706).

Visit website

Outbound clicks are tokenized for directory analytics.