Prompt Security

Prompt Security is an AI-security company focused on protecting generative-AI usage at runtime across the enterprise. Its platform gives organizations real-time visibility into how employees and applications use AI tools, what data is being shared, and where risk arises, then applies policy-based controls that block high-risk prompts, prevent sensitive-data leakage, and stop prompt-injection and jailbreak attacks. The platform is model-agnostic, covering major providers including OpenAI, Anthropic, and Google, and secures three surfaces: employee use of third-party GenAI (a GenAI firewall / DLP for shadow-AI), homegrown GenAI applications, and AI agents, including MCP-gateway security that mediates connections to thousands of known MCP servers. Founded around 2023 and based in Tel Aviv, Israel, Prompt Security raised roughly $23M from investors including Hetz Ventures, Jump Capital, Ridge Ventures, Okta, and F5. In August 2025, SentinelOne agreed to acquire the company (reportedly around $250M) to extend its AI-native Singularity platform into GenAI and agent security. Prompt Security is aimed at security and risk teams that need both governance over sanctioned and unsanctioned AI use and inline runtime protection for the AI applications and agents they build.