Hardware wallet maker Ledger has launched the Ledger Agent Stack, a new open-source toolkit designed to let AI agents interact with cryptocurrency wallets without ever gaining direct control over users' assets. The announcement reflects the growing convergence of artificial intelligence and blockchain, while maintaining Ledger's long-standing focus on security.
The toolkit allows AI agents to check wallet balances, prepare swaps, draft transactions, and suggest blockchain operations, but every action involving funds must still receive explicit approval on a physical Ledger hardware wallet before execution.
Ledger believes AI assistants will become increasingly common across the crypto industry but argues that human verification should remain mandatory for sensitive operations. As the company explains, it is preparing for a future where AI agents "become more autonomous" while deliberately taking a cautious approach to protect users from mistakes and security risks.
Security Comes Before Automation
Unlike some emerging crypto wallets that allow AI agents to execute transactions automatically, Ledger's system keeps final control firmly in the hands of users. Every signing request remains secured by the company's hardware devices, preventing compromised software or malicious AI agents from moving funds independently.
Ledger highlighted growing concerns surrounding AI security, noting that "the AI attack surface is growing faster than the defense." According to the company, 26.1% of AI agent skills contain at least one security vulnerability, while approximately 60% of security breaches result from human error.
The company believes combining AI automation with mandatory human confirmation significantly reduces both types of risk.
AI Is Expanding Across the Crypto Industry
Ledger's announcement comes as AI-powered crypto assistants become increasingly common. Wallet providers and blockchain platforms are racing to integrate intelligent agents capable of simplifying decentralized finance, trading, and asset management.
One notable example is MetaMask, which is also developing agentic capabilities while adopting a similar human-in-the-loop security model that requires users to approve transactions before execution.
Rather than competing on automation alone, Ledger is positioning its solution around security-first AI, allowing users to benefit from AI assistance without sacrificing custody of their private keys.
The Ledger Agent Stack also includes open-source Markdown instructions that allow developers to integrate Ledger hardware directly into AI workflows. According to the company, MoonPay and Shisa.ai have already begun incorporating these capabilities into their products.
Beyond Crypto Wallets
Ledger also sees its hardware becoming a broader digital security platform. Thanks to support for the OpenPGP encryption standard, developers can use Ledger devices to protect API keys, AI credentials, and other sensitive authentication data outside the cryptocurrency ecosystem.
The company says its hardware can also function similarly to YubiKey security devices, allowing users to securely authenticate accounts across platforms including GitHub, npm, 1Password, Discord, and other online services.
The launch forms part of Ledger's 2026 AI Roadmap, which includes additional features such as AI agent identity management, secure execution environments, and "proof of human" verification scheduled for release in the coming months.
Founded in 2014, the Paris-based company continues to expand beyond hardware wallets while reportedly exploring an initial public offering (IPO) that could value the business at more than $4 billion. As AI adoption accelerates across the digital asset industry, Ledger is betting that secure human oversight will remain an essential component of the next generation of crypto infrastructure.
