Runlayer, Rippling drop suits – a founder cautionary tale

▼ Summary
– Runlayer and Rippling dropped their lawsuits against each other without a settlement or money exchanged, and Rippling immediately released its MCP gateway product at the center of the dispute.
– Runlayer, a startup with $42 million in funding led by Andrew Berman, sued Rippling after a year of product testing, claiming Rippling cloned its MCP gateway without becoming a customer.
– An MCP gateway securely manages enterprise AI agent data requests from other software systems, adding features like role-based access control and usage logs.
– Rippling countersued Runlayer for patent violations, a move Runlayer saw as an attempt to increase legal costs and force it to drop its suit.
– The legal fight highlights that in the fast-changing AI landscape, startups face unpredictable competitors, as Rippling—historically a payroll firm—has now entered the AI gateway and security market.
Two startups spent weeks trading legal blows, only to walk away with nothing settled. On Wednesday night, Runlayer and Rippling simultaneously dropped their lawsuits against each other. No settlement, no cash exchanged, and according to court documents viewed by TechCrunch, not even attorney fees were paid out.
Rippling didn’t waste a moment celebrating. It immediately released its MCP gateway, the exact product at the center of the dispute and the one that directly competes with Runlayer’s own offering.
For founders, this episode reads like a modern fable. In the AI era, where spinning up new software is almost effortless, your next rival might not come from a competitor’s ranks. It could be a customer who kicked the tires on your product, then decided to build their own version.
Here’s the short version of the legal spat. Runlayer is a young startup that came out of stealth in November 2025, having raised $42 million from backers including Khosla Ventures’ Keith Rabois and Felicis. Its leader is Andrew Berman, a third-time founder whose prior ventures include Nanit, a baby-monitor company, and Vowel, an AI video conferencing tool that Zapier acquired in 2024.
According to Runlayer’s complaint, Rippling tested its MCP gateway for over a year, with engineering teams from both companies working side by side. But Rippling never became a paying customer. Instead, Berman got a text from a Rippling employee saying the company was building its own MCP gateway and planned to ship it as a product. That same employee reportedly called Rippling’s version a clone of Runlayer’s.
Runlayer sued, arguing that Rippling broke contractual agreements tied to the product testing.
For context, an MCP gateway is a security layer that manages an enterprise’s AI agent requests for data from other software systems. Say a recruiter asks for the top five candidates for a role, including their emails. That data sits in the company’s hiring platform. The gateway retrieves it on behalf of the agent, instead of handing agents direct access to internal systems. It can also add features like role-based access control, so managers see different data than interns, plus observability tools that log usage and create audit trails.
Rippling fired back with a countersuit, accusing Runlayer of patent infringement. Runlayer viewed that move as a pressure tactic designed to inflate legal costs and force a retreat.
After three weeks of discovery, Runlayer dropped its case. Rippling dropped its own, with no settlement collected on either side.
So the lawsuits fizzled into a very public, very expensive spat. But there’s a broader lesson buried in the noise. The AI landscape is shifting so quickly that the long technical evaluations enterprises love to run on startups need a serious rethink. Between the day an AI startup enters one of those trials and the day it ends, months later, the customer’s needs can change beyond recognition.
Meanwhile, Rippling, a company long known for payroll and benefits, has pivoted into the AI gateway market in a matter of weeks. Its tool routes requests to different models and tracks token spending per employee. That puts it in direct competition with Stripe, Ramp, and Databricks.
Rippling has also entered the AI security space, using its MCP gateway to tie AI access to employee roles. That puts it up against Runlayer, Docker, and Amazon Bedrock.
As for Runlayer, it’s pitching a broader package of agent security services wrapped around its gateway. That includes agent creation and the ability to spot shadow AI agents running inside an enterprise without IT’s knowledge. The fight may be over, but the market they’re both chasing is only getting more crowded.
(Source: TechCrunch)




