Outerlimit

Decentralized zero-trust security layer that authorizes and binds AI agent actions at the moment of tool execution.

Outerlimit · Security and safety tooling · security

Published in Feed

Overview

Outerlimit emerged from stealth on September 22, 2026 with a decentralized authorization layer that extends Zero Trust principles to the "agent action layer" -- the point where AI agents actually execute tools and take actions, rather than just the identity or network perimeter. The platform works in three stages: Discover (mapping agents, tools, and shadow AI across an environment), Observe (provable visibility into what agents are doing), and Enforce (deterministic, cryptographic controls that bind identity, authorization, and action together at the moment of execution, with credentials fragmented across the agent ecosystem to remove single points of compromise). Outerlimit raised a $16 million pre-seed round led by AlbionVC, Evolution Equity Partners, and Crane Venture Partners, with strategic angel backing from executives at Alan, ReliaQuest, A-lign, Goldman Sachs, FTV Capital, Twin Track Ventures, Garrison, and GP Bullhound. Founded by CEO Tony Pepper, Neil Larkins, and CTO Dr. Peter Vincent (a theoretical neuroscientist), the company targets Fortune 500 and FTSE 100 enterprises deploying agentic AI at scale.

This is a sourced directory entry. Features and capabilities are reported in the linked sources; this entry does not document independent Aymorphic testing.

Key features

  • Three-stage model: Discover, Observe, Enforce across the agent action layer
  • Cryptographic enforcement binding identity, authorization, and action at the moment of tool execution
  • Fragmented, distributed credential and key storage across the agent ecosystem to remove single points of compromise
  • Multi-hop chain integrity preservation across agent-to-agent and agent-to-tool call chains
  • Shadow AI and unmanaged agent/tool discovery across the enterprise

Use cases

  • Securing autonomous AI agents that have access to enterprise systems and data
  • Detecting and blocking rogue or compromised agent behavior in real time
  • Extending existing enterprise zero-trust programs to agentic AI deployments

Pricing

unknown -- pricing not disclosed at launch; go-to-market is targeted at Fortune 500 and FTSE 100 enterprises

Pricing reflects the source record and may change. Confirm current terms with the provider.

Dates and sources

Released
22 Sept 2026
Record updated
30 Sept 2026
Record last checked
30 Sept 2026

Explore the Feed

Other entries filed under security and safety tooling; inclusion does not imply equivalent features.

  • DefenseClaw Edge Connector

    Cisco's open-source DefenseClaw gains an Edge Connector that allows or blocks AI agent tool calls directly on IoT and robot hardware.

  • Lakera Guard 2.0

    Lakera's real-time guardrail API for LLM apps, updated with multi-agent protection and a new benchmark for detecting indirect prompt injection in tool outputs.

  • Fleming-1

    Sierra's model that listens to a live phone call and scores whether the caller is an AI agent rather than a human.

Browse the Feed