Ledger Targets AI Agent Risks With Hardware-Based Security and Human Controls

Hardware that helps protect you from rogue AI agents will soon be launched by French crypto security firm, Ledger, which announced it’s set to enter the artificial intelligence (AI) security space by releasing a range of new technologies.

In a blog post written by the firm’s Chief Human Agency Officer,  Ian C. Rogers, published April 14, Ledger outlined products designed to ensure AI agents remain secure and their behaviour aligned with human intentions. 

Rogers argued that software-based security is insufficient given AI agents’ access to sensitive information and their increasingly powerful capabilities. “Ledger is launching a comprehensive security stack for AI Agents throughout 2026,” he said.

While AI agents need access to money, credentials, and identity to be useful, software-only security is insufficient for production-grade risk.

Ian C. Rogers, Ledger

Ledger said it would strengthen AI agents’ trustworthiness by introducing “hardware-anchored security to the agentic economy.” The new devices would function much like Ledger’s existing crypto hardware solutions, storing sensitive information on the device and requiring physical button presses to execute certain tasks.

Ledger’s suite of AI agent security technologies to roll out in 2026 will include: 

• Device management kit (DMK) which has already been released, and is being used by Moonpay. 
• Hardware-anchored identity and wallet for AI agents, along with an agent command line interface, will roll out in Q2.
• Agents Intents, which Ledger refers to as a “human-in-the-loop approval layer” for agents, to roll out in Q3.
• Agent Policies, allowing humans to use the hardware device enforce rules on agents (such as sending limits), coming in Q4.
• Proof of Human, allowing the human behind AI agents to prove their identity, due in Q4.

Related: Japanese Giants Unite to Build the Brain Behind the World’s Machines

Why Is Hardware Based Security Necessary for AI Agents?

Rogers warned that “an agentic future is coming, but an agentic future where we give agents our logins, credit cards, and identities is a security nightmare.”

As Ledger sees it, providing our most sensitive information to powerful AI agents on our behalf is a risky proposition where outcomes could be very bad for individuals if those agents take action the human didn’t intend.

Related: Ledger Integrates OKX DEX to Enable In-App Multichain Token Swaps

Rogers said Ledger’s fundamental purpose as a business derives from its belief that “digital private property is real, and you have the right to own and control it.” This belief has informed their approach to crypto security and it’s now the basis of their entry into AI security.

“Your agents will hold your API keys, your credentials, your identity, and your money,” he said.

We believe ownership and control must be grounded in hardware. A secure element doesn’t care if the surrounding software is compromised. The signing boundary still holds. Human approval still holds.

Ian C. Rogers, Ledger