Key takeaways
- Ledger has confirmed an unauthorized hardware implant in affected devices, with reported losses now exceeding $86 million.
- This is a different and more serious mechanism than the pre-seeded recovery phrase that early coverage, including ours, described.
- A pre-seeded device announces itself: a genuine wallet never arrives with a recovery phrase. A hardware implant can let the device behave entirely normally.
- It belongs to the interdiction class of attack, where hardware is modified between manufacture and delivery — long documented against high-value targets, rarely seen at this scale in consumer products.
- Firmware attestation checks authenticity of software, so depending on where an implant sits it may pass those checks. Provenance becomes the control that matters.
Ledger has confirmed that affected devices contained an unauthorized hardware implant. Reported losses now exceed $86 million, and the cases continue to centre on units obtained through a reseller rather than from the manufacturer.
This is worth a second article rather than a footnote, because it is a materially different attack from the one most coverage described first — including ours yesterday, which set out the pre-seeded recovery phrase mechanism. That is the common version of this attack. It is not what happened here.
The difference between the two attacks
The pre-seeded attack is a trick played on the user. The device arrives with a recovery phrase already decided, usually printed on a card, presented as yours. It works because people follow the instruction in the box.
Its weakness, from the attacker's point of view, is that it is visible. A genuine hardware wallet never arrives with a recovery phrase. Anyone who knows that one fact is immune, and the card in the box is the evidence.
A hardware implant removes that tell. A physical component added inside the device can compromise key generation or capture the secret at the moment it exists, while the device continues to behave exactly as expected. You generate your own phrase, on the device, on its own screen, and the protection you believed that gave you is not there.
So the advice that defeats the first attack — generate the phrase yourself, never accept a supplied one — does not defeat the second. That is the uncomfortable part.
This attack class has a name
Modifying hardware between manufacture and delivery is called interdiction, and it is well documented. It has historically been the business of state intelligence agencies against high-value targets, because it requires physical access to the specific units and real technical capability.
What is unusual here is the setting. Interdiction against consumer products, at this scale, through a retail channel, is rare. It is also a rational economic decision by an attacker: a device that will hold a large balance justifies far more effort than a phishing email does, and the attacker chooses which devices to prepare.
That reframes the risk. The question is not whether you are individually interesting enough to target. It is whether the channel you bought through was.
Why the genuineness check may not catch it
Manufacturers provide a device authenticity check through their own software, and it is a real control worth running. It is also worth understanding what it verifies.
These checks generally confirm that the secure element holds authentic manufacturer keys and that the firmware is genuine and unmodified. That addresses firmware tampering and cloned devices well. Depending on where an implant sits and what it does, it may leave the attested components untouched and still achieve its purpose.
So run the check. Do not treat passing it as proof that the hardware is unmodified, because it was not designed to answer that question.
What this leaves you with
Being honest about this means admitting the defences are narrower than the usual advice implies.
Provenance does the work. Buying directly from the manufacturer, or through a channel the manufacturer authorises, is the control that addresses this at source, and it is the one thing every reported case appears to have in common by its absence. Second-hand and marketplace resellers are not a discount on the same product; they are a different risk profile.
Beyond that: a small test transaction, then waiting, then a larger one, costs almost nothing and would have surfaced several of these cases early. Splitting holdings across devices from different sources limits how much any single compromised unit can cost. And for larger amounts, multisignature arrangements using devices from different manufacturers mean one compromised device is not sufficient to move funds — which is the only arrangement that structurally survives this attack.
The wider point, again
Yesterday's conclusion holds and this strengthens it. Self-custody does not remove trust; it relocates it. Holding your own keys removes an exchange's solvency and competence from the equation and substitutes the manufacturer, the firmware, the randomness of the chip, and the physical path the device travelled to reach you.
Most of those you can reason about. The last one is the one almost nobody checks, and it is the one that failed here.
Education, not investment advice.
Frequently asked questions
What is a hardware implant in this context?
A physical component added inside the device at some point between manufacture and delivery. Depending on its design it can compromise key generation or capture the secret when it exists, while the device otherwise behaves normally.
How is this different from a pre-seeded wallet?
A pre-seeded device arrives with a recovery phrase already decided, often on a card in the box. That is visible — a genuine wallet never supplies a phrase. An implant has no such tell: you generate your own phrase on the device and the protection you expect is simply absent.
Does the manufacturer's genuineness check detect it?
Not necessarily. Those checks generally verify that the secure element holds authentic keys and the firmware is genuine and unmodified, which addresses firmware tampering and cloned devices. An implant may leave the attested components untouched. Run the check, but do not read passing it as proof the hardware is unmodified.
Is interdiction a new kind of attack?
No. Modifying hardware between manufacture and delivery is well documented and has historically been used by state intelligence agencies against high-value targets, because it needs physical access and real capability. What is unusual is seeing it against consumer products at this scale through a retail channel.
What actually protects against it?
Provenance, primarily — buying from the manufacturer or an authorised channel, which is the factor the reported cases share by its absence. Beyond that: a small test transaction before committing funds, splitting holdings across devices from different sources, and for larger amounts multisignature across different manufacturers, so one compromised device cannot move funds alone.
Should I stop using hardware wallets?
That is not the conclusion. A hardware wallet from a verified source still removes a large set of risks that holding keys on a general-purpose computer or at an exchange does not. What this case changes is how much weight the purchase channel deserves relative to everything else.
Not financial advice. Crypto assets are volatile and unregulated in many jurisdictions. In India, gains are taxed at 30% with 1% TDS on transfers. Do your own research and never invest money you cannot afford to lose.
Editorial note: Crypto Shakti uses an AI-assisted research and drafting workflow. Every article is grounded in the linked primary sources and live market data captured at publication time.