Ledger Agent Stack is a new open-source development platform designed to help developers build AI-powered cryptocurrency agents without compromising security, allowing artificial intelligence to automate wallet management while requiring every transaction to be approved by a human using a Ledger hardware device.

Announced as the first major milestone in Ledger’s 2026 AI roadmap, the platform extends the company’s hardware wallet security model into the growing field of AI-powered financial automation. Rather than giving AI direct control of digital assets, Ledger has built Agent Stack around a hardware-enforced trust model where AI agents can prepare transactions, but users retain ultimate authority over every signed action.

The approach is designed to let AI assistants monitor wallet balances, prepare transfers, recommend token swaps, analyse portfolios and automate routine tasks without exposing private keys or allowing autonomous access to cryptocurrency funds.

Hardware Security for the AI Era

Ledger Agent Stack brings the security architecture used across millions of Ledger hardware wallets into an open-source toolkit for developers building AI applications.

AI agents can perform read-only operations such as checking balances, transaction history and wallet information independently. However, any operation requiring a digital signature—including transfers, swaps, staking or other on-chain transactions—must be physically approved on a Ledger hardware wallet.

Ledger is also expanding hardware-backed protection beyond cryptocurrency through two new applications designed to secure developer identities and credentials.

The new OpenPGP application encrypts agent secrets so they remain inaccessible unless a Ledger device is connected, while the Security Key application enables hardware-backed authentication for services including GitHub, npm, Discord and 1Password. According to the company, the additions help address growing risks associated with compromised developer credentials and increasingly automated software environments.

Four Open-Source Components

Ledger Agent Stack consists of four core open-source tools aimed at developers, enterprises and institutional cryptocurrency users.

The Device Management Kit (DMK) Skills provides Markdown-based instruction sets that allow large language models and AI frameworks to integrate directly with Ledger hardware. Ledger says organisations including MoonPay and Shisa.ai have already implemented DMK support.

The Ledger Wallet CLI gives AI agents programmable access to wallet information, balances, receiving addresses and transaction preparation for transfers, swaps and staking. While informational tasks can be completed automatically, every signing request still requires confirmation on a Ledger hardware wallet.

For enterprise customers, the Ledger Enterprise CLI applies the same security principles to institutional treasury management through hardware security module (HSM)-backed policy enforcement and governed transaction workflows.

The Ledger Enterprise Multisig CLI extends those capabilities to enterprise multisignature wallets, allowing AI agents to prepare, query, sign and execute transactions while maintaining mandatory human oversight throughout the approval process.

Addressing AI Security Challenges

As AI agents become increasingly capable of managing financial workflows, Ledger argues that security must remain anchored in dedicated hardware rather than software alone.

The company points to research suggesting human error contributes to around 60 per cent of security breaches, while more than one-quarter of AI agent skills contain at least one security vulnerability.

By ensuring every signing operation requires physical confirmation through trusted hardware, Ledger says Agent Stack significantly reduces the likelihood of compromised AI environments authorising unauthorised cryptocurrency transactions.

Part of Ledger’s Broader AI Strategy

Ledger Agent Stack is the first public release under the company’s broader AI strategy announced earlier this year.

Future roadmap initiatives include Agent Identity, Agent Intents and Policies, and Proof of Human, all designed to strengthen hardware-enforced trust for AI-powered applications throughout 2026.

Ledger says its long-term vision is to establish hardware security as the foundation for autonomous digital systems, allowing AI to assist users without replacing human oversight of critical financial decisions.

Ian Rogers, Chief Human Agency Officer at Ledger, said the company’s wallets already support one of the industry’s broadest cryptocurrency ecosystems.

“Ledger has the most full-featured suite of wallets. The most chains, providers for buy, swap, and yield. Also, Ledger has wallets for Enterprise and Multisig users. Now your agent is your trusted partner in operating these secure wallets.

“Crypto wallets have protected billions on this standard for years. Ledger Agent Stack allows your agent to use these wallets just as easily as humans.”

Ledger says the release aligns with CEO Pascal Gauthier’s vision that hardware should remain the ultimate root of trust as AI becomes increasingly embedded in digital finance, ensuring people remain in control of every critical transaction.

Why this matters

The launch of Ledger Agent Stack reflects a broader shift across the cryptocurrency industry as developers explore AI-powered wallet management, trading assistants and autonomous financial agents. By combining artificial intelligence with hardware-enforced human approval, Ledger is positioning its platform to address growing concerns around AI security, digital asset protection and enterprise governance while enabling developers to build more capable cryptocurrency applications.