Kraken rebuilds its mobile app around an AI investing assistant with manual trade approval
The exchange says the system monitors markets and recommends trades and portfolios tied to user goals, but will not execute autonomously.
Kraken said it is redesigning its mobile app around an AI-driven investing assistant that tailors recommendations to user goals and preferences. The exchange framed the system as decision support that requires user approval for every trade rather than autonomous execution.
Key Takeaways
- Kraken is redesigning its mobile app around AI that personalizes the interface and investing recommendations to user-set goals and preferences.
- The exchange’s “financial intelligence” is built to monitor markets continuously and surface trade ideas, while keeping execution manual through required user approval.
- Goal-based flows like buying a home, saving for retirement, and building an emergency fund are central to the redesigned experience.
- Chief data officer Kamo Asatryan said the aim is to give everyday users the same market awareness as Kraken’s most active traders, using plain-English prompts as the interface.
Kraken’s AI-First Mobile App Redesign Keeps Trading Execution in the User’s Hands
Kraken published details of a redesigned mobile app centered on an AI investing assistant it describes as “financial intelligence.” The company’s core positioning is explicit: the system continuously monitors markets, identifies investment opportunities, and recommends trades, but it does not place transactions on its own.
That constraint matters for market structure. By requiring user approval before any trade is placed, Kraken is drawing a bright line between decision support and hands-off automation. It is also a practical way to reduce the blast radius of model error in fast markets, even as the assistant is marketed as always-on.
Kraken also framed the redesign as part of a broader push into financial services, with personalized investing tools meant to sit above basic trading features.
From Order Tickets to Goal-Based Portfolios: What Kraken Says the Assistant Will Do
The redesigned workflow starts with intent, not instruments. Kraken said users begin by setting financial goals and preferences, and the app then tailors its interface and recommendations around those objectives rather than pushing customers through complex trading tools.
The company described goal-based use cases including buying a home, saving for retirement, and building an emergency fund. In practice, that signals a shift in UX away from a tools-first order-ticket experience and toward portfolio construction and ongoing guidance, which could broaden the product’s appeal beyond the most active traders.
Kraken said the assistant uses a user’s goals alongside risk tolerance, funding preferences, and financial profile to generate a suggested portfolio that the user can review and adjust before investing. After a user is invested, the app is designed to provide personalized portfolio updates and investment suggestions tailored to the user’s holdings.
Kraken chief data officer Kamo Asatryan tied the product to continuous market awareness. He said the technology is meant to give everyday investors the same awareness as Kraken’s most active traders, adding: "[T]here's an opportunity for everyday people to become high-frequency traders and do so using plain English,".
AI Agents Are Spreading Across Exchanges and Fintech: OKX, Coinbase, and Revolut X
Kraken’s approach lands in the middle of a widening autonomy spectrum. OKX launched a beta marketplace in June 2026 where AI agents can transact autonomously, complete onchain tasks, and build blockchain-based reputations.
Coinbase also introduced a tool in June 2026 that lets AI agents make payments and trade cryptocurrencies on behalf of users using its x402 payments protocol. Chainalysis reported “last month” that agentic payment activity on Coinbase’s Base network surpassed 100 million transactions, with transaction growth stabilizing while higher-value transfers became more common.
Revolut, by contrast, shipped an upgrade to Revolut X that allows customers to connect AI assistants including Claude, Gemini, Cursor, and OpenClaw to analyze markets, backtest strategies, and place orders via natural-language prompts. Like Kraken’s design, that flow requires users to review and approve each trade before execution.
Details Kraken Hasn’t Shared Yet: Launch Timing, Coverage, and Risk Controls
Kraken has not provided a launch date or rollout schedule for the redesigned app, and it has not clarified whether availability will differ by region or by mobile platform.
The exchange also has not specified which products the assistant will cover, including whether recommendations apply only to spot markets or extend into derivatives or other offerings. For traders, that scope determines whether the assistant is a retail onboarding layer or a tool that could meaningfully change how risk is expressed.
Kraken has not published performance metrics, backtesting results, or details on risk controls for the recommendation engine, including how it behaves in fast markets. Competitive pressure is also a live variable: if rivals expand from “approval required” into more autonomous execution, Kraken will have to decide whether to hold the line on manual approval or compete on automation.
The Retail-Flow Implication of ‘Recommendations Only’ AI Trading
I read Kraken’s design choice as a deliberate attempt to capture the upside of AI-driven engagement without taking on the headline risk of autonomous execution. The threshold that matters is whether the assistant can consistently generate recommendations that users actually act on, because “approval required” still concentrates the outcome in conversion and retention, not in theoretical model capability.
The real test is whether Kraken pairs goal-based UX with credible risk controls and product coverage that matches how users trade. If manual approval holds while recommendations expand across more markets, the setup starts to look structural rather than narrative-driven, because it would reshape retail flow through a simpler interface instead of outsourcing execution to an agent.
