Runlayer Sues Rippling Over Alleged MCP Gateway Copying

The delicate boundary between corporate evaluation and product replication is under scrutiny once again in the artificial intelligence ecosystem. According to AI News & Artificial Intelligence | TechCrunch, startup Runlayer has filed a lawsuit against HR and workforce management platform Rippling, alleging that the incumbent built its own Model Context Protocol (MCP) gateway after reviewing Runlayer's proprietary technology.

While the legal specifics will unfold in court, the dispute highlights a recurring hazard for early-stage infrastructure providers: showing their architecture to well-capitalised enterprise platforms under the guise of partnership or vendor assessment.

The Perils of Enterprise Evaluation

The Model Context Protocol has quickly emerged as an important standard for connecting large language models to external tools, databases, and enterprise applications. As organisations move from experimental AI pilots to production deployments, managing these connections securely requires robust gateway infrastructure. Startups like Runlayer have moved rapidly to fill this void, designing specialised tooling to handle authentication, routing, and telemetry for agentic workflows.

Navigating early-stage enterprise partnerships requires startups to balance commercial exposure against the existential threat of incumbent replication.

However, enterprise software platforms frequently face a classic build-versus-buy decision when new architectural patterns gain traction. When an established platform engages with a niche startup, the line between standard market research and intellectual property infringement can easily blur. For early-stage companies, demonstrating a product's technical superiority to a potential enterprise client often requires exposing core design choices and system architecture.

Protecting IP in the Age of Open Standards

This litigation underscores broader structural tensions within the AI infrastructure landscape. Because protocols like MCP are designed around open, interoperable standards, the competitive advantage for gateway providers lies not in the underlying protocol itself, but in the operational middleware—governance, speed, security, and developer experience.

When incumbents evaluate these specialised gateways, they are often attempting to assess whether the functionality belongs as a feature within their broader suite rather than a standalone vendor contract. If a platform decides to build an in-house alternative after inspecting an early-stage product, proving unlawful appropriation requires demonstrating that specific trade secrets or proprietary code were misappropriated, rather than merely copying an architectural concept.

Strategic Implications for Infrastructure Founders

For founders building AI infrastructure, this dispute offers a stark reminder of the operational risks inherent in corporate discussions. To safeguard their assets while pursuing commercial traction, emerging vendors must carefully structure their engagement processes.

First, technical deep dives during vendor evaluations should be strictly compartmentalised. Demonstrating value through sandbox environments or restricted pilots can limit the exposure of underlying architecture. Second, robust non-disclosure agreements with explicit terms around derivative works and evaluation boundaries remain essential, even if enforceability varies by jurisdiction.

Ultimately, as enterprise demand for AI agent integration accelerates, incumbent software providers will continue seeking integrated control over context management. How the court handles Runlayer’s claims against Rippling will provide critical guidance on where product inspiration ends and intellectual property theft begins in the fast-evolving enterprise AI market.