Longbridge extends AI strategy beyond its own app, letting outside assistants trade on its platform
The AMW Read
Longbridge's MCP/API rollout follows an already-established pattern set by Interactive Brokers and Moomoo, so it confirms a known trajectory with impact confined to Asia-focused retail brokerage.
Longbridge extends AI strategy beyond its own app, letting outside assistants trade on its platform
Longbridge Securities has introduced developer tools — its API, the Model Context Protocol (MCP), and predefined AI "skills" — that let external AI assistants including ChatGPT, Claude, Gemini, Cursor, and Codex access brokerage functions such as market data retrieval, portfolio monitoring, news analysis, and trade preparation. The rollout, framed around an "Invest Loop" workflow spanning market observation, analysis, trade planning, order placement, and post-trade review, sits alongside LongbridgeAI, the company's existing in-app assistant. Longbridge says users remain responsible for investment decisions and trade confirmations while it handles connectivity and order placement. The launch coincides with regional expansion, including a new Singapore headquarters at Asia Square, alongside existing Hong Kong and US market access covering stocks, ETFs, REITs, warrants, and options.
The move places Longbridge in a widening group of brokerages exposing account and trading functions directly to general-purpose AI assistants rather than confining AI to a proprietary chat layer. Interactive Brokers has already connected ChatGPT, Claude, and Grok to client accounts for portfolio analysis and trade preparation, and Moomoo (Futu Holdings) has shipped comparable API-based capabilities, following earlier in-app tools such as Tiger Brokers' TigerGPT. As commissions, onboarding speed, and app features converge across brokerages, exposing infrastructure through MCP and skills is becoming one of the few remaining differentiators, moving the integration point from a brokerage-controlled interface to whichever assistant a customer already uses.
For builders, MCP-based brokerage access signals that financial-account connectivity is turning into a standard "skill" surface for general-purpose assistants, extending use cases beyond research summarization into order preparation and portfolio monitoring. For investors evaluating this space, differentiation is shifting from pricing and UI polish toward which brokerages expose the deepest, safest API surface to third-party AI — a governance and authorization question that remains unresolved industry-wide, since safeguards around data access, accuracy, and regulatory compliance are still early and largely untested at scale.