The AI Wallet Problem: Why Security Hardware Makers Are Drawing a Line on Agent Autonomy
Ledger and others are warning that AI agents with unsupervised access to digital wallets represent a fundamentally different — and more dangerous — threat model than anything crypto security was designed for.
As AI agents become more capable, the question of what they should be allowed to do with money is becoming urgent. Hardware wallet maker Ledger fired a shot across the bow this weekend, as @Ledger posted: "imagine giving your ai agent permission to do whatever with your digital assets without your physical approval — that's gonna be a hard pass from us." The statement is notable not for its content — physical approval for transactions is Ledger's entire business model — but for the fact that a major security company felt the need to say it publicly.
Unlock the full briefing
Get every story in today's briefing, the full archive, and the daily AI intelligence brief.
All stories today
Full archive
Daily brief
Cancel anytime. Payments powered by Stripe.