Key facts

Founded
2018
HQ
New York, United States
Employees
—
Website
Funding stage
Series B · US$42m
Further facts
  • Funding: Arthur raised a US$42m Series B in September 2022, co-led by Acrew Capital and Greycroft with Index Ventures, BAM Elevate, Work-Bench and Plexo Capital.
  • Filing: Arthur filed an SEC Form D in August 2026 for an equity offering first sold in July 2026; the filing records US$2.67m sold to eight investors against an offering size of up to US$6m.
  • Timeline: Arthur open-sourced its real-time evaluation engine in March 2025, launched the Agent Discovery and Governance platform in December 2025 and listed it on Google Cloud Marketplace in January 2026.

Focus areas

Agent governance LLM evaluation AI observability AI runtime security AI-SPM

About

Arthur is a New York-based AI governance and evaluation company founded in 2018 by chief executive Adam Wenchel. Its platform gives security, risk and AI engineering teams a central view of every model and agent running in the enterprise, with continuous evaluation, runtime guardrails, audit trails and cost controls. Customers include Axios, Expel and Upsolve. Arthur raised a US$42m Series B co-led by Acrew Capital and Greycroft, and its Agent Discovery and Governance platform is sold through Google Cloud Marketplace.

How they use AI

Arthur's Agent Discovery and Governance platform scans endpoints, clouds and on-premises systems for AI agents, including unregistered shadow agents, then traces each agent's reasoning steps and tool calls. Continuous evaluations score groundedness, tool selection, relevance and PII leakage, runtime guardrails enforce acceptable-use policies, and behavioural baselines flag anomalies such as data-egress spikes, with findings streamed to CrowdStrike Falcon, Splunk, Datadog and Elastic. Routing sends requests that do not need an expensive model to cheaper or internally hosted open-source models, and the open-source Arthur Engine runs real-time evaluations inside a customer's own environment.

Products and services

  • Agent Discovery and Governance (ADG) — Finds and governs every AI agent across endpoints, cloud and on-premises
  • Arthur Evals — Pre-production and runtime evaluation for LLM apps and agents
  • Arthur Runtime Security — Guardrails and behavioural anomaly alerting for agents
  • Arthur Engine — Open-source real-time AI evaluation engine
  • Cost control — Token spend tracking and routing to cheaper models

Notable customers / partners

Key people

  • Adam Wenchel — Co-Founder and Chief Executive Officer

Investors