Arrakis Security emerges from stealth to secure the AI workforce
AI agents are moving from assistants to autonomous operators — and enterprises are racing to secure systems that can make decisions, access sensitive data and act on their behalf
Arrakis security co-founders. Photo: Arrakis website
Israeli cybersecurity startup Arrakis Security has emerged from stealth with an $8 million seed round and a platform designed to secure the rise of autonomous AI agents inside enterprises.
“AI agents have become the new digital workforce,” wrote Tal Baron, co-founder and CEO of Arrakis, in a blog post announcing the company’s launch. Unlike traditional chatbots, these systems are increasingly capable of making decisions, executing workflows and accessing critical business applications.
But as enterprises rapidly adopt AI agents, a new security gap is emerging. Baron calls it the “visibility illusion”: organizations may know that an AI agent exists and what systems it can access, but often lack insight into what the agent is actually doing, which permissions it is using and when legitimate behavior becomes a security risk.
“You simply cannot secure a machine that thinks and operates differently from a human using security tools designed to monitor humans interacting with machines,” Baron wrote.
Founded by Baron, Omer Efrat and Ron Shani, former members of Torq and Palantir, Arrakis is building a Behavioral Governance platform designed to discover AI agents, create behavioral profiles, monitor their actions and detect risky activity across enterprise environments.
The company is also launching Arrakis Labs, a research division focused on AI agent security threats. The team includes former Microsoft security researchers and veterans of elite military technology units.
The $8 million seed round was led by Hetz Ventures, with backing from cybersecurity and AI executives including Mati Staniszewski, CEO of ElevenLabs; Ofer Smadari, CEO of Torq; Amitai Ratzon, CEO of Pentera; and senior executives from Palantir.
According to Hetz Ventures General Partner Pavel Livshiz, the security layer around AI agents is still in its early stages, and enterprises need ways to govern these systems without slowing AI adoption.
With the new funding, Arrakis plans to expand its R&D team in Israel as companies move from experimenting with AI assistants to deploying autonomous systems across business operations.
In an unusual coincidence, another AI startup named Arrakis also emerged publicly this week – but the two companies appear to be unrelated. While the Israeli Arrakis Security is focused on protecting AI agents, the London-based Arrakis is building and deploying AI agents for industrial companies, helping organizations in sectors such as aerospace, energy and manufacturing automate mission-critical workflows.
The shared name is a nod (most probably intentional) to the fictional desert planet from Frank Herbert’s Dune, the only source of the universe’s most valuable resource: spice. As AI becomes the new strategic resource for enterprises, it seems more than one startup has drawn inspiration from the same futuristic universe.