Arcade.dev has secured $60 million in Series A funding to develop a secure action layer designed to move AI agents from pilot phases into production. The round, led by SYN Ventures with strategic investments from Morgan Stanley and Wipro, aims to address the critical authorization and governance gaps that currently prevent Fortune 500 companies from deploying agents at scale. While many enterprises are racing to put agents into production, most remain stuck in pilot because security teams cannot verify which agent took which action, on behalf of which user, and against which system.
SYN Ventures Leads $60M Series A for Arcade.dev
The $60 million Series A funding brings Arcade's total capital to $72 million, following a $12 million seed round in 2025. As part of the investment, Jay Leek, Managing Partner at SYN Ventures, has joined the company's board. Leek noted that agents are currently hitting a "wall" where adoption has outrun the infrastructure required to make them safe, positioning Arcade as a company built for production reality from day one.
The company intends to use the new capital to accelerate hiring, ecosystem growth, and product development as enterprises scale agent deployments from small pilots to thousands of production workflows. The founding team brings deep expertise to this challenge, having previously developed the identity, data, and integration layers that became industry standards for organizations including Okta, Redis, MongoDB, Snowflake, and Airbyte. This technical foundation has driven rapid growth, with the company reporting that tool call volume has increased 25x over the last six months.
Arcade's Action Layer and MCP Authorization Specification
Arcade focuses on the "action layer," distinguishing its technology from MCP gateways. While a gateway merely routes traffic, Arcade's layer authorizes, executes, and governs the action itself. The company authored the MCP authorization specification adopted by Anthropic and is currently operating in production for LangChain, Prosus, and a top US bank. According to the company, it has shipped more tools for delegated user authorization than the rest of the ecosystem combined.
The platform addresses three specific barriers to production deployment:
- Authorization: The system ensures agents receive the same access as the user for a specific action. By eliminating standing permissions and overprivileged service accounts, Arcade prevents leaked PII or data loss and limits the blast radius of agent hallucinations.
- Reliability: Arcade provides over 8,000 MCP tools purpose-built for agent usage rather than using standard API wrappers. This specialized approach results in fewer failed actions and lower token costs for the enterprise.
- Governance: The infrastructure maintains a complete audit trail identifying which agent performed an action, on behalf of which user, and against which specific resource.
Enterprise Deployment and Governance Requirements
The funding comes as Fortune 500 security teams struggle to track agent actions and user permissions across various systems. Co-founder and CEO Alex Salazar stated that agents often fail in production not because of model errors, but because organizations cannot prove that an agent is authorized to perform a specific action on a resource on behalf of a user.
Zheng Wang, Head of Strategic Investments at Morgan Stanley, noted that the infrastructure is designed to ensure proper authorization and governance as enterprises integrate AI agents across their operations. By providing the necessary infrastructure to prove authorization, Arcade aims to be the layer through which every serious enterprise agent deployment runs.
Key Takeaways
- Arcade.dev raised $60 million in a Series A led by SYN Ventures, bringing total funding to $72 million.
- The company authored the MCP authorization specification adopted by Anthropic and provides over 8,000 purpose-built MCP tools.
- The platform is currently used in production by LangChain, Prosus, and a top US bank to manage agent authorization and audit trails.
TechInsyte's Take
Arcade is positioning itself as the critical governance layer for the "action" phase of AI agent workflows, moving beyond simple routing to strict authorization. For CIOs and CISOs, the primary value lies in the elimination of overprivileged service accounts, though the long-term scalability of the MCP authorization specification across diverse legacy enterprise stacks remains to be seen. Decision-makers should monitor how this infrastructure integrates with existing identity providers to ensure a seamless audit trail.
Source: Businesswire
