Enigma Launches With $71M to Build Foundation Models for Robots

Israeli startup Enigma has emerged from stealth with $71 million in funding to develop large‑scale “foundation models” tailored for robots. The company aims to provide a general‑purpose intelligence layer that can run across a wide range of machines, from household devices to industrial robots, instead of each manufacturer building its own narrow AI stack.

Enigma’s founding team reportedly includes alumni of Israel’s intelligence community and engineers from a cybersecurity startup that was acquired by Google, bringing together deep expertise in both robotics and applied AI. Foundation models for robots are designed to fuse vision, sensor, and language data so machines can navigate complex, changing real‑world environments with less task‑specific programming.

If this approach succeeds, it could significantly cut the cost and lead time for deploying robots in logistics, manufacturing, and even elder care. But it will also sharpen questions for regulators: how to certify safety when autonomy is driven by a shared model, and who is responsible when an AI‑controlled robot makes a mistake—the model provider, the robot maker, or the operator?