Key Insights
- Kraken crypto added personalized portfolios and approval-based trade recommendations.
- Robinhood allowed connected agents to execute trades without individual approval.
- Coinbase and Revolut also expanded artificial intelligence trading access.
Kraken crypto is rebuilding its mobile app around artificial intelligence tools that personalize portfolios and recommend trades. The exchange announced the redesign on July 10, targeting users who found conventional trading interfaces difficult to manage.
The product placed Kraken inside a widening contest over AI trading and agent-controlled finance. Robinhood, Coinbase, and Revolut had already introduced tools that let software agents analyze markets or place orders.
Kraken Crypto App Adds Goal-Based Investing
Kraken said users would first enter financial goals, preferences, and existing holdings. The app would then adapt its interface and investment plan around those inputs.
Kraken listed home purchases, retirement savings, and emergency funds among possible objectives. Its system would monitor markets and identify actions aligned with each user’s plan.
The exchange described the feature as persistent financial intelligence rather than a one-time portfolio generator, CNBC reported. Kraken support materials said it could flag idle cash, recurring purchases, or positions requiring rebalancing.
Users would still decide whether any suggested transaction reached the market. Kraken said its system would never place a trade without direct approval.
That control separated the product from fully agentic accounts. It also limited execution risk when market conditions moved faster than a user could review recommendations.
Kraken Keeps Users Above Every Trade
Kraken chief data officer Kamo Asatryan told CNBC that the system targeted broader access to continuous market monitoring. He said ordinary users could eventually trade through plain-English instructions.
However, Kraken’s design stopped short of granting autonomous execution rights. The app analyzed a financial profile, proposed a portfolio, and delivered updates after investment.
That model resembled an advisory interface layered over an exchange account. It did not turn the customer’s capital over to an independent software agent.
Kraken’s approach also reflected earlier work around automated strategies. The company acquired Capitalise.ai in August 2025 to support no-code strategy testing and execution through Kraken Pro.
In March, Kraken released a command-line interface for artificial intelligence agents. The tool included paper trading against live market data before users deployed capital.
Robinhood Pushes AI Trading Further
Robinhood launched Agentic Trading on May 27 and expanded its plans on July 1. The company said customers could connect third-party agents to dedicated accounts for automated investment decisions.
Robinhood stated that agents could execute orders without approval for every transaction. Its disclosure warned that rapid AI-driven strategies could become difficult to monitor or stop.
The company later said eligible U.S. customers would gain similar agent connections for crypto accounts. That proposal moved Robinhood beyond Kraken’s recommendation-and-approval structure.
Robinhood had also introduced Cortex, an artificial intelligence research assistant for market analysis. The broader product strategy combined research, account controls, and autonomous execution inside one financial platform.
The contrast created separate risk models for customers. Kraken kept execution authority with users, while Robinhood delegated defined trading decisions to connected agents.
Robinhood’s model offered faster execution, but it exposed customers to broader automation risk. Kraken’s approval layer reduced that exposure, though it also introduced delays during volatile markets.
For less active investors, the distinction could shape product choice. Some users may prefer guided decisions, while others may accept automation for faster portfolio management.
Kraken Crypto Faces a Crowded Agent Market
Coinbase launched Agents trading on June 11 through command-line and Model Context Protocol integrations. The company said agents could trade, pay, and run account workflows within user-defined limits.
Revolut also connected Revolut X with Claude, Gemini, Cursor, and OpenClaw. Its June release allowed natural-language analysis and order preparation, while customers approved final trades.
These products showed exchanges moving beyond charts, order books, and custody services. They increasingly competed through personalization, conversational access, and automated account actions.
The competition also shifted attention toward interface design and user trust. Exchanges now had to explain how agents accessed accounts, applied limits, and handled execution errors.
The systems also raised different questions around consent and oversight. Recommendation tools preserved direct approval, while autonomous agents transferred more operational control to software.
Kraken had not announced a public launch date beyond saying the redesigned app was coming soon. Its rollout will test whether guided portfolios can compete with Robinhood’s autonomous crypto trading model.
