Interactive Brokers opens MCP standard AI connectivity for all clients

IBKR extends its AI trading integration to any MCP-compatible tool, marking a structural shift in how retail and institutional clients interact with financial data.

By Disrupts Editorial Team Thu, 13 August 2026

A modern conference room features a long dark wooden table with a central dark grey inset panel and power modules, surrounded by light grey upholstered chairs with metal frames, all brightly lit by natural light from large windows showcasin

Interactive Brokers (Nasdaq: IBKR) has opened its AI trading integration to any application built on the Model Context Protocol (MCP), the open standard that has rapidly become the de facto connector layer between AI assistants and external data services. Previously restricted to certified marketplaces for ChatGPT, Claude, and Grok, the integration now supports a broader universe of MCP-compatible tools including Claude Code, Cursor, Perplexity, and Windsurf, with Google Gemini compatibility described as forthcoming.

The move is operationally straightforward: clients add the IBKR connector URL to any MCP-compatible tool's custom server settings, authenticate through Interactive Brokers' secure login, and authorise account access. From there, they can query their portfolios, model risk scenarios, and draft trade instructions in natural language. Critically, in this release, all AI-drafted instructions require manual client review and conversion to an order before execution, meaning the human remains in the loop at the point of trade submission.

Brokerage infrastructure meets the agentic AI layer

The timing reflects a broader pattern in fintech infrastructure. MCP, originally introduced by Anthropic and since adopted widely, has emerged as the connective tissue linking AI assistants to live data sources, much as REST APIs standardised web service communication in the previous decade. Interactive Brokers has positioned itself early in this architecture: the company says its AI integration was built on MCP from launch, making today's announcement an expansion of scope rather than a platform rebuild.

Milan Galik, Chief Executive Officer of Interactive Brokers, framed the move within the company's longer-standing API-openness philosophy: "Supporting AI tools built on the Model Context Protocol is a natural extension of that philosophy. It gives clients the flexibility to use the AI applications they prefer to research investments, manage their portfolios, and interact with their IBKR accounts."

The feature set extends beyond portfolio querying. IBKR's existing AI suite, accessible directly within its own platforms, already includes natural-language stock screeners across more than 70,000 global equities, thematic search tools, and AI-generated news summaries filtered to a client's specific holdings. The MCP integration now means a professional investor could, in principle, wire those capabilities into a custom-built agentic workflow running in a coding environment such as Cursor or Windsurf, without waiting for a certified marketplace approval.

Second-order implications for wealth tech and agentic capital

For the cross-sector strategist, the more consequential signal here is structural: Interactive Brokers is effectively turning its brokerage data layer into a plug-and-play module for whatever AI orchestration stack a sophisticated client prefers to run. That dynamic has significant implications beyond retail finance.

Hedge funds and proprietary trading desks that are already building internal AI agents on MCP-compatible frameworks can now pull live account data, risk metrics, and implied volatility figures directly into those workflows. This erodes one of the traditional moats of prime brokerage: the proprietary data terminal. As open standards commoditise the connectivity layer, competitive advantage shifts up the stack toward model quality, signal generation, and execution speed rather than data access itself.

The move also sits within a wider capital and talent realignment. Wealth management platforms, robo-advisers, and independent financial adviser networks have been watching agentic AI deployment closely but cautiously, wary of regulatory exposure around autonomous order placement. IBKR's human-in-the-loop constraint on this release is a deliberate positioning choice that may serve as a template for competitors navigating the same compliance tension between automation ambition and client-protection obligations in the UK, EU, and Singapore markets the firm serves.

As MCP adoption widens and AI orchestration matures, the next competitive question for the brokerage sector is not whether to connect to AI tools, but how quickly to permit those tools to act, not merely advise.