What Does IBKR's Certified Claude and ChatGPT Connector Actually See — and What Is the Chart Context Gap?

What Is the IBKR Certified MCP Connector?

In June 2026, Interactive Brokers launched a certified Model Context Protocol (MCP) connector for Claude (Anthropic) and ChatGPT (OpenAI). It is listed in both platforms' connector marketplaces as a certified integration. No API keys, no developer account, and no code are required — IBKR clients connect through the standard account authentication flow.

MCP is an open standard that lets AI assistants connect to external data sources and tools. IBKR's implementation follows this standard, meaning any MCP-compatible AI can connect to the same data surface. The certification label indicates IBKR has verified the integration against each platform's requirements.

At launch, the connector supports equities and ETFs with market and limit orders. IBKR indicated additional asset classes would follow. For current coverage and the setup guide, see the IBKR Traders' Insight page on connecting Claude and ChatGPT to an IBKR account.

• IBKR press release — June 1, 2026

  Official Interactive Brokers announcement of the certified Claude and ChatGPT connector.

• IBKR Traders' Insight: How to Connect Claude AI and ChatGPT to Your IBKR Account

  Step-by-step connector setup guide from IBKR.

What Data Does the Connector Expose to Claude and ChatGPT?

The connector gives the AI access to: account summary (balances, net liquidation value, buying power), open positions with cost basis and current values, margin utilization, trade history, and market data — both real-time quotes and historical price data for equities and ETFs.

Crucially, AI-generated orders do not auto-execute. When Claude or ChatGPT drafts a trade instruction, it routes to a dedicated 'AI Instructions' tab inside the IBKR client interface. The trader reviews and manually submits. This is the compliance architecture: the AI proposes, the human decides.

The data surface is meaningful. An AI with your positions, cost basis, current prices, and historical quotes can answer questions about portfolio state, flag concentration risk, and draft structured order instructions for review in the AI Instructions tab.

What the Connector Cannot See: The Chart Context Gap

The connector knows what you own, at what price, and what the market is doing. It does not know why you own it. The chart context — the trendlines drawn before entry, the support zones marked, the indicator confluence that validated the setup — is invisible to the AI through the connector.

This is the chart context gap. The AI can see you are long 500 shares at a certain price with current P&L. It cannot see that your entry was a support-zone retest at a prior structure level, that your exit rule fires if price breaks below a specific level, or that the setup belongs to a pattern you have been tracking across three sessions.

For a technical trader, that missing layer is most of the trade logic. Market data is public. Position data is mechanical. The annotated chart state — the level notes, the drawing geometry, the session tags — is the proprietary context that distinguishes a structured process from a raw account view. For what that context layer looks like when captured explicitly, see How to Make IBKR Chart Work AI-Readable.

Why the Gap Matters in Practice

If you ask Claude 'what should I do with my position?' through the connector, the AI reasons from price, cost basis, and any publicly available information about the instrument. It does not reason from your chart analysis — your analysis is not in the connector's data scope.

That is not a defect in the connector. It is a scope decision: the connector is an account and market data interface, not a charting system. The gap becomes a problem only when traders expect the connector to substitute for chart-context-aware AI workflow. It cannot. The connector answers account questions. Chart questions require chart context.

Traders who want AI to reason about their annotated charts — not just their account — need to export that context explicitly. For how IBKR chart data, levels, and drawing notes can be made AI-readable, see Claude Code IBKR Chart Data Workflow and How to Turn IBKR Chart Notes Into an AI-Readable Trading Journal.

What the Connector Is Genuinely Useful For

The connector is well-suited to portfolio-level account questions: position summaries across symbols, sector concentration review, P&L checks, buying power calculations, and drafting simple equity or ETF orders for review in the AI Instructions tab. For this class of tasks the connector works without supplemental chart context.

The practical limit appears at any question that requires knowing why a position was taken, what the exit conditions are, or whether the current chart setup remains valid. Those questions require chart-context export alongside the connector — not instead of it.

A complete AI-assisted workflow combines the connector for brokerage-side data with a structured chart annotation export for technical context. For the architecture of a human-reviewed workflow that combines both layers, see How to Build a Human-Reviewed AI Workflow Around IBKR Charts. For the signals-vs-process question the connector raises, see Can Claude or ChatGPT Give You Trade Signals Through the IBKR MCP Connector?.

What does IBKR's certified connector give Claude and ChatGPT access to?

Account balances, open positions with cost basis, margin status, trade history, and market data including real-time and historical quotes for equities and ETFs. Asset class coverage expands over time from the June 2026 equities-and-ETFs launch.

Does the IBKR connector include chart drawings or technical indicator state?

No. The connector exposes account and market data only. Chart drawings, annotated support and resistance levels, marked zones, indicator values, and session review notes are outside the connector's data scope.

Can Claude or ChatGPT execute trades automatically through the IBKR connector?

No. AI-generated trade instructions go to an 'AI Instructions' review tab in the IBKR client interface. The trader must review and manually submit. Nothing executes without explicit client action.

What is the chart context gap with the IBKR connector?

The gap is the difference between what the AI can see through the connector — account state and market prices — and what it cannot see — chart drawings, annotated levels, TA indicator state, and setup notes. Most of the reasoning behind a technical trade lives in chart context, not account data.