Kraken Rebuilds Its App Around AI Trading as Exchange Race Accelerates
Kraken is relaunching its mobile app from the ground up with agentic AI at the core, targeting retail investors ahead of an anticipated IPO, while simultaneously releasing an open-source developer tool that allows AI agents to interface directly with its markets. The consumer app requires explicit human confirmation before any trade executes; the command-line tool is built for developers and AI agents that may operate with greater autonomy.
Kraken announced this week that its forthcoming mobile app will not be a feature update to its existing product. The exchange stated plainly that the new version would not be "an AI assistant or a copilot bolted onto the old version of the app." Instead, the AI logic is embedded from the first screen: during onboarding, the system assesses a user's goals, risk tolerance, funding preferences, and financial profile, then generates a draft portfolio with an explanation of its reasoning.
Users approve or modify before any funds move. After that initial setup, the AI monitors the portfolio, surfaces curated news, and flags idle cash for potential redeployment.
All trade recommendations still require the user to confirm before execution. The system is described as "agentic" in the sense that it monitors conditions and proactively surfaces opportunities, not in the sense that it acts without human sign-off.
Kraken's Chief Data Officer, Kamo Asatryan, framed the product in terms of access. "It's important for customers who are more everyday people to have that same capability and be as well informed as the professional traders," he said.
The Developer Layer
Alongside the consumer app, Kraken has open-sourced a command-line interface (CLI) built in Rust and aimed at developers and AI agents rather than retail users. The tool covers 134 commands across spot trading, futures, staking, and transfers. It outputs data in NDJSON format, a line-by-line structure that machines parse more reliably than standard formatted text, per Kraken's own documentation.
It also includes a built-in Model Context Protocol (MCP) server. MCP was originally developed by Anthropic, and its inclusion here means AI coding environments including Claude Code, OpenAI Codex, and Cursor can connect to Kraken's markets without developers writing custom exchange adapters. A local paper trading engine allows teams to test agent strategies against live market data without risking capital.
The rationale from Kraken's own blog was blunt: "APIs are designed for systems. But when you ask an AI agent to interact with a raw crypto exchange API, it struggles."
Fourth to Market, Not First
Kraken is the fourth major exchange to ship agentic trading features. Bitget launched its Agent Hub in March 2026. Gemini followed in late April, integrating directly with both Claude and ChatGPT via MCP and marketing the launch as "the future of crypto is autonomous." Coinbase released its own agentic trading tools in June, though its implementation leans more toward stablecoin payment infrastructure than active trade execution. Of the three US-regulated exchanges in this group, Kraken's approach is the most consumer-facing, embedding AI logic into the core onboarding flow rather than routing it through a separate developer interface.
AI-driven bots now account for approximately 58 percent of all crypto trading volume globally, according to crypto industry analysis. Gartner projects that 40 percent of enterprises will have adopted agentic AI workflows by the end of 2026. Between May 2025 and April 2026, AI agents settled roughly $73 million across 176 million blockchain transactions, with an average transaction size of $0.48, per Hashed Emergent analysis. McKinsey estimates that AI agents could mediate $3 to $5 trillion in consumer commerce by 2030, a figure that frames the current exchange-level tooling as early infrastructure for a much larger shift.
Coinbase CEO Brian Armstrong has stated publicly: "Very soon there are going to be more AI agents than humans making transactions." Not everyone agrees on the current state of the technology. Haseeb Qureshi, managing partner at Dragonfly Capital, offered a more cautious read: "Everything here is basically a toy right now."
Regional Implications
The product has real relevance outside the US, though the picture is uneven. Kraken supports users in Nigeria, Kenya, Pakistan, and India, among other markets. In Nigeria specifically, the exchange does not integrate the naira, meaning users there face foreign exchange conversion friction on every deposit and withdrawal.
Pakistan presents a distinct opportunity. The country has one of the world's largest unbanked populations and a substantial diaspora remittance market, making AI-driven portfolio tools particularly relevant as a crypto-native savings instrument for users who lack access to conventional financial products.
For developers, the CLI is the more immediately useful output. Any team using Claude Code or Codex in Lagos, Nairobi, Karachi, or Bangalore can now interface with Kraken's markets through the MCP server without building custom connectors. The paper trading engine lowers the barrier further for bootstrapped teams that cannot afford to test live with real capital.
Africa's largest crypto exchange by volume, VALR, launched its own agent-compatible AI service in April 2026, suggesting the infrastructure trend is developing in parallel on the continent rather than simply flowing from US platforms. That thesis has further support at the startup layer: NjiaPay, a pan-African agent-based payment startup, raised $2.1 million in March 2026, building infrastructure for the same category of agentic financial interaction.
India presents a specific complication. The Reserve Bank of India has explicitly rejected private stablecoins, which creates friction for any agentic workflow that settles through stablecoin layers. A concrete local precedent is already taking shape: in February 2026, Razorpay and NPCI piloted agentic payments via Anthropic's Claude, enabling conversational ordering on Zomato, Swiggy, and Zepto using the same MCP architecture that underpins Kraken's CLI. Kraken has stated ambitions to build a full-stack financial services platform spanning payments, banking, and lending. How it navigates the RBI's stance in a large retail market like India will matter.
Regulatory pressure is tightening globally as well. The EU AI Act's high-risk provisions become enforceable on August 2, 2026, setting a compliance benchmark against which AI-led trading products will increasingly be measured, regardless of where those products are headquartered.
What Comes Next
The product launch is also a capital markets story. Kraken filed a draft S-1 confidentially in November 2025, initially targeting a Q1 2026 listing at an $800 million raise against a $20 billion valuation. It paused those plans in March 2026 amid market turbulence. Deutsche Börse invested $200 million in April 2026, implying a current valuation of approximately $13.3 billion. A Q3 2026 IPO is now expected.
With 5.7 million funded accounts globally, Kraken enters that process with a consumer base large enough to anchor a retail-facing product narrative. An AI-first mobile app gives the company a concrete story to carry into that process, at a moment when the case for retail crypto platforms needs to look like more than just a trading fee business.