Runlayer, Rippling drop lawsuits — but the brouhaha is still a cautionary tale for founders

The high-stakes legal standoff between startup Runlayer and enterprise software giant Rippling has come to an abrupt end. On Wednesday evening, both parties voluntarily dismissed their respective lawsuits. According to court filings, the resolution was clean: no settlement was reached, no financial compensation was exchanged, and neither side covered the other’s legal expenses.

However, the conclusion of the litigation was far from a quiet affair. Rippling immediately signaled its intent to dominate the space by launching its own MCP (Model Context Protocol) gateway—the very technology that served as the catalyst for the legal battle and a direct competitor to Runlayer’s core offering.

A Cautionary Tale for the AI Era

For the broader startup ecosystem, this public spat serves as a stark reminder of the shifting power dynamics in the age of artificial intelligence. As building sophisticated software becomes increasingly accessible, the barrier to entry has lowered, turning potential partners into immediate rivals overnight.

"In the age of AI, when building new software has become almost trivial, you never know who your next competitor will be. It might even be a prospective customer."

The conflict highlights a growing vulnerability for early-stage companies: the risk of sharing proprietary technical insights with larger enterprises during long, drawn-out testing phases.

The Anatomy of the Dispute

Runlayer, an early-stage venture that emerged from stealth in November 2025, has secured $42 million in funding from prominent backers, including Keith Rabois of Khosla Ventures and Felicis. Led by serial entrepreneur Andrew Berman—whose track record includes the sale of Vowel to Zapier—the company positioned itself as a leader in AI agent security.

The friction began when Rippling spent over a year testing Runlayer’s MCP gateway. According to Runlayer’s initial complaint, the two engineering teams collaborated closely, only for the relationship to sour when Berman received word from a Rippling staffer that the company was building a "clone" of the Runlayer product.

Runlayer sued, alleging a breach of contractual agreements established during the product testing phase. Rippling responded with a countersuit citing patent infringement—a move widely interpreted as a tactical maneuver to drain Runlayer’s resources. After three weeks of discovery, both sides opted to walk away.

What is an MCP Gateway?

At its core, an MCP gateway acts as a secure intermediary for enterprise AI agents. Rather than granting AI direct access to sensitive internal systems, the gateway:

• Retrieves Data: Safely fetches information from software systems (e.g., pulling candidate data from recruitment platforms).
• Enforces Security: Implements role-based access control (RBAC), ensuring interns and managers have different levels of data visibility.
• Provides Observability: Tracks logs and usage trails to ensure AI agents are operating within corporate guardrails.

The Future Landscape

The rapid evolution of this market suggests that the traditional "technical shoot-out" model—where enterprises force startups into lengthy evaluation periods—is becoming obsolete. By the time a startup completes a pilot, an enterprise’s requirements may have shifted entirely.

Rippling is clearly pivoting toward a more aggressive stance in the AI infrastructure market. Beyond its payroll and benefits roots, the company is now competing with heavyweights like Stripe, Ramp, and Databricks in the AI routing space, while simultaneously challenging Runlayer, Docker, and Amazon Bedrock in the AI security sector.

Runlayer, meanwhile, continues to differentiate itself by offering a broader suite of agent security services, including tools to detect "shadow AI" agents operating without IT oversight. While the courtroom drama has concluded, the competitive race to secure the enterprise AI stack is only just beginning.