WitnessAI Raises $58 Million for Global Expansion and Announces New Ways to Secure AI Agents

Strategic Funding Round, Led by Sound Ventures and Joined by Corporate Investors Samsung Ventures and Qualcomm Ventures, Advances the Company’s AI Security Capabilities and Reach

WitnessAI, the AI security platform trusted by leading enterprises, announced strategic funding of $58 million. Led by Sound Ventures, an early investor in OpenAI, Anthropic, and SentinelOne, and with participation from Fin Capital, Qualcomm Ventures, Samsung Ventures, and Forgepoint Capital Partners, the funding will be used to accelerate WitnessAI’s global go-to-market and product expansion. Today, the company also unveiled expanded agentic AI governance capabilities that bring observability to global enterprises developing and deploying AI agents.

“We are the only AI security vendor that can secure every AI interaction, everywhere, with a unified solution.”

Also Read: CIO Influence Interview with Duncan Greatwood, CEO at Xage Security

Expanded Agentic AI Security Capabilities

WitnessAI provides the confidence layer for enterprise AI adoption. The leading AI security platform extends to secure agents in two novel ways. The first is securing AI agent activity similarly to LLM activity, monitoring which agents are active, what MCP servers and tools they are accessing, and what data they are sharing. WitnessAI can see what an agent is doing on an employee’s or customer’s behalf. By connecting the human and agentic identities and capturing the evolving decision-making context, including agent state and execution commands at runtime, WitnessAI provides explainability for agent actions. This creates powerful observability across both workforces, human and agentic.

The second is extending AI application protection from AI models to agents, safely blocking attacks and malicious prompts before they reach the agent. These capabilities enable the protection of multi-generational AI apps, built on foundational model APIs, custom LLMs, and AI agents. The platform uniquely provides policy control based on behavioral intent, understanding the meaning and intention of any prompt. This enables smarter, more accurate policy enforcement, effectively blocking advanced threats such as prompt injection or multi-turn attacks.