Pandorex
Security

Exein Raises $270 Million for Embedded Security in Robots and Drones

Published Pandorex Redaktion·2 min read
—
Illustration: a red-protected AI chip blocks attacks on a robot arm, drone and industrial controller.
Editorial illustration · Pandorex

In brief: Italian security company Exein has raised $270 million at a $1.7 billion valuation. The capital will expand kernel-level protection and autonomous defence for robots, drones, vehicles and industrial equipment. The promised AI models have not yet been independently validated.

Defence inside the machine

Headline led the round, joined by Sofina, Goldman Sachs, EIB Group through the European Tech Champions Initiative, KfW Capital and T.Capital. Exein is also expanding a revolving credit facility led by J.P. Morgan. It plans US and Asian growth plus possible acquisitions in Europe and the US.

Exein's existing Photon runtime architecture operates at kernel level and is intended to stop malicious execution before impact. The company is now training a proprietary foundation model on two years of machine telemetry from a footprint exceeding two billion devices. Autonomous security agents based on it are meant to react without waiting for human intervention.

The agentic architecture is due by the end of 2026 and the first foundation models in the first quarter of 2027. Exein reports roughly 5,000 new, non-repetitive attacks per week, five times last year's volume. These are company figures; it has not disclosed the methodology, device mix or attack definition.

Pandorex Analysis

The approach targets a real gap: an attack on a robot or controller can alter physical actions, not merely steal data. Runtime protection shortens response times compared with a patch-only model and can cover equipment whose firmware is rarely updated.

Autonomous defence also moves risk into the security layer. A false positive can stop machinery, while a missed attack can cause physical harm. False-positive rates, resource use, update and rollback paths, and agent isolation will therefore matter. Funding and telemetry scale demonstrate reach, not reliable defensive effectiveness.

Sources and references

Sources used for the facts and context in this article.

  1. Exein, 15.09.2026: Funding and Physical AI security roadmapexein.io
  2. Financial Times, 15.09.2026ft.com

How Pandorex researches and corrects articles

Comments

Sign in to write a comment.

Swipe up
Next Article

Report: Palantir Demands Zero Retention as Anthropic Excludes Frontier Models

Security