The Swiss-regulated bank outlined the three initiatives in a Sept. 7 announcement, describing them collectively as a major expansion of its digital offering. Dukascopy says it serves more than 400,000 clients worldwide and is regulated by the Swiss Financial Market Supervisory Authority.
The most technically notable element is the bank’s integration of artificial intelligence into its JForex trading infrastructure through the Model Context Protocol, or MCP.
The system allows clients to connect AI assistants including ChatGPT and Claude to their JForex trading accounts. Through natural-language commands, traders can execute orders, manage existing positions, analyze markets, calculate risk parameters and monitor account exposure without navigating a conventional trading terminal or writing code.
Dukascopy says the integration does not require programming knowledge and can be configured within minutes. Client passwords are not shared with the AI assistant. The service is currently available for JForex demo accounts, while support for live accounts is expected to follow.
The announcement represents the latest stage of a rollout that began earlier in the year. Dukascopy first announced its AI-powered trading service on June 23, describing itself as the first Swiss bank to launch AI-powered trading through MCP.
The broker’s expansion of stock trading also predates the Sept. 7 announcement. On June 5, Dukascopy launched a dedicated Stock Trading Platform offering more than 25,000 CFDs on stocks and ETFs across 20 markets, compared with more than 1,500 stocks and ETFs previously available through its JForex environment. The company says the number of markets could eventually expand to 87.
The stock platform is structured separately from the existing JForex account. It is accessed through JForex4 Desktop and uses a dedicated stock CFD sub-account. Unlike the broader JForex offering, the new stock environment currently supports long positions only and offers 1:1 leverage. Dukascopy says automated strategies are not supported on the stock platform.
The third component is Dukascopy’s new mobile banking application.
The app replaces the bank’s Connect 911 and Swiss Mobile Bank applications and combines banking, payments, cards, foreign exchange and investment services in one interface. Customers can remotely open accounts through video identification, manage payment cards, make international transfers and access investment services from their mobile devices. Dukascopy initially announced the flagship app on June 19.
Taken together, the initiatives show Dukascopy attempting to broaden its identity beyond a conventional forex broker. Its existing platform already covers forex, stocks, cryptocurrencies, indices and gold, while the bank also provides multicurrency accounts and IBAN services.
The timing is notable because the broker industry is increasingly competing on technology as well as pricing and instrument count. AI assistants are moving from research tools toward interfaces capable of interacting directly with financial accounts, while brokers are simultaneously expanding access to equities and other asset classes.
For Dukascopy, the challenge will be turning those individual launches into a coherent ecosystem rather than simply adding more features to an already broad product lineup.
AI Trading Moves From Analysis Toward Execution
The most consequential part of Dukascopy’s expansion is arguably not the 25,000 CFDs. It is the decision to put an AI assistant between the customer and the trading account.
For years, brokers have used AI for relatively low-risk functions such as market summaries, sentiment analysis, customer service and trading ideas. Allowing an AI interface to interact with an account moves the technology much closer to execution.
That distinction matters.
A chatbot that tells a trader that EUR/USD has strengthened is fundamentally different from one that can receive a natural-language instruction, calculate an order and send that order to a brokerage account. The latter creates a new layer between human intent and financial execution.
Dukascopy’s use of MCP is therefore potentially more important than the marketing label of “AI-powered trading.” MCP provides a standardized way for AI applications to interact with external tools and services. Connecting that framework to a brokerage account could eventually allow traders to build much more sophisticated workflows without learning traditional programming.
But it also introduces questions that have little to do with whether the AI can understand a trading instruction.
What happens when an ambiguous command produces an unintended order? How are position limits enforced? How does the system distinguish analysis from authorization to trade? What happens if the model misunderstands the instrument, quantity or direction? And who is responsible when an AI-assisted instruction produces a loss?
Those questions become especially relevant because AI systems are probabilistic while financial transactions are not. A misunderstood sentence in a normal chatbot conversation can be corrected in the next message. A misunderstood trading instruction can become a real position within seconds.
Dukascopy’s decision to initially make the integration available on demo accounts may therefore be significant. It gives the bank an opportunity to test the interaction model before allowing customers to connect AI directly to live capital. The eventual live rollout will be the more meaningful milestone.
The 25,000-CFD expansion tells a different story. It is primarily about breadth. Moving from roughly 1,500 stocks and ETFs to more than 25,000 gives Dukascopy a much larger equity universe, although the current 1:1 leverage and long-only structure mean it should not be interpreted as simply replicating a conventional leveraged CFD platform.
The banking app is arguably the glue connecting these developments. If customers can bank, fund accounts, trade multiple asset classes and eventually execute trades through AI from the same ecosystem, Dukascopy can compete on convenience rather than only on spreads or leverage.
That may ultimately be the real strategy behind the three launches.
The broker is trying to reduce the distance between banking, investing and trading. The AI interface could become the front door, the mobile application the everyday financial hub and the expanded CFD universe the product breadth behind it.
Whether customers actually want that degree of integration remains the open question. But if AI-assisted execution becomes a mainstream feature of retail trading, Dukascopy has positioned itself unusually early in that transition.
The strategy also echoes a broader move toward portfolio management and execution infrastructure becoming increasingly automated. For brokers, the competitive advantage may increasingly come from how seamlessly these different functions work together rather than from any single feature.