Snapchat Adds MCP Server for AI Ad Campaign Management

Snapchat Joins the MCP Wave

Snapchat has become the latest platform to embrace the Model Context Protocol (MCP), rolling out a server that lets brands connect their advertising campaigns directly to AI agents such as Claude, ChatGPT and Gemini 1. The move allows marketers to manage promotional efforts on Snapchat through conversational AI tools rather than navigating the platform's native ad dashboard, reflecting a broader push across the advertising and media industries to make business data and workflows accessible to autonomous AI agents.

Part of a Broader Industry Pattern

Snapchat's announcement is not an isolated event but part of a rapidly accelerating trend of companies building MCP servers to bridge their platforms with popular AI assistants. Real estate software provider Rechat recently launched its own MCP server, enabling agents to use ChatGPT and Claude to manage contacts, deals and marketing campaigns tied to their business data 2. Reuters has taken a similar approach on the media side, introducing an MCP server that gives customers programmatic access to its news content for use in agentic workflows, positioning trusted journalism as a resource AI systems can pull into their outputs 3. Marketing analytics firm Oviond has also joined the movement, launching an MCP server that lets agencies handle client reporting through AI tools including ChatGPT, Claude and Cursor 4.

Taken together, these launches illustrate how MCP has quickly become a standard connective layer between enterprise software and large language models, letting agents retrieve data, execute tasks and automate workflows across advertising, real estate, media and marketing analytics without custom-built integrations for each AI provider.

Security Concerns Cast a Shadow

Even as adoption spreads, the MCP ecosystem is facing mounting scrutiny over security. One report describes a significant disclosure of unpatched vulnerabilities in the MCP standard, warning that the protocol's role as a connective layer between AI models and external tools makes any weaknesses especially consequential 5. Separate coverage points to more than 40 CVEs logged in just four months, arguing that as large language models move from experimental use into mission-critical enterprise infrastructure, MCP security has evolved from a technical concern into a board-level issue 6.

Why It Matters

The simultaneous rise of MCP adoption and MCP security warnings highlights a tension at the center of the current AI agent boom: businesses are racing to plug their platforms into ChatGPT, Claude, Gemini and other assistants to stay competitive, but the underlying protocol enabling this connectivity is still maturing. As companies like Snapchat, Rechat, Reuters and Oviond expose sensitive business data and campaign controls to autonomous agents, the vulnerabilities flagged in recent security reports suggest that governance, auditing and patching practices will need to keep pace with the speed of integration, or organizations risk exposing critical systems through the very tools meant to streamline them.