Bug Found In BitBox Hardware Wallet

Bug in BitBox Hardware Wallet

Share this article

On August 17, 2026, Swiss hardware wallet maker BitBox disclosed and patched two “severe” vulnerabilities in its BitBox02 and BitBox02 Nova devices, alongside a previously flagged bootloader issue. It’s the second major hardware wallet security story in as many months, arriving just weeks after COLDCARD’s catastrophic RNG flaw was linked to over $112 million in stolen Bitcoin, thus far.

Unlike COLDCARD, BitBox reports they’ve found and patched these bugs in time, and no funds have been reported as lost or stolen, with no exploitation detected in the wild.

Under normal circumstances, a vendor coming out this firmly with news and a patch would be applauded, but given the fact that we’ve seen the biggest cold storage breach in history, no amount of PR and goodwill will settle the nerves.

The only thing that can do that is seeing your funds reflected when you boot up a device and check your addresses.

What Was Actually Found?

BitBox’s Dixence update (firmware version 9.26.5) fixed two distinct issues, plus a bootloader flaw already patched in an earlier release.

The memory corruption flaw

Affected Multi editions of the BitBox02 and BitBox02 Nova, but only in a specific state: before the device has been set up with a wallet, while connected to a malicious host computer. Basically, all new BitBox users were at risk.

Under these conditions, a compromised host could trigger memory corruption on the device, potentially enabling arbitrary code execution — meaning an attacker could push malicious firmware onto the device before the legitimate owner ever generated a seed.

Note: BitBox’s Bitcoin-only edition was not affected, since its firmware doesn’t contain the vulnerable code path at all, a direct benefit of its intentionally smaller codebase. Yet another case of why being Bitcoin-only can save you!

The Silent Payments flaw

This was arguably the more novel of the two.

Silent Payments is a Bitcoin privacy feature that lets a recipient receive funds without publishing a fresh address for every transaction.

The bug meant a malicious host device could manipulate a Silent Payments transaction so that funds got locked to an address the sender didn’t actually intend.

Critically, BitBox clarified this wasn’t a direct theft vector — an attacker couldn’t simply redirect coins into their own wallet — but it opened the door to a ransom-style attack, where a victim’s Bitcoin becomes inaccessible unless the attacker cooperates to help recover it.

The bootloader flaw

While not part of this round of fixes, it is worth mentioning since the patch went in this year. Fixed in the July “Oeschinen” update and reiterated in this disclosure, it concerned the code responsible for deciding what firmware a device is allowed to accept.

A flawed bootloader is dangerous because it undermines the entire chain of trust a hardware wallet depends on — if attackers can get unauthorised firmware validated as legitimate, everything downstream (transaction signing, key handling, display verification) can no longer be trusted either.

Notably, BitBox said its own internal audits — conducted with help from frontier AI models — caught all three issues before any public disclosure of exploitation, and the company reports zero confirmed cases of stolen funds tied to any of them.

Since the cat is out of the bag with AI, it’s really become a game of white hat versus black hat prompt engineering, and I think this has to become best practice: have all frontier models brute-force any codebase and patch updates from now on.

How This Could Theoretically Lead to Stolen Funds?

It’s worth being precise about the actual attack chain here: yes, it was a bug, and yes, funds could have been stolen, but the attack vector was much smaller. COLDCARD’s flaw was a passive weakness baked into already-generated seeds — funds could be at risk even without the owner doing anything wrong, simply because the randomness behind their key was flawed.

BitBox’s flaws required active exploitation conditions: a malicious or compromised host computer, and in the memory corruption case, a device that hadn’t yet been initialised with a wallet.

That’s a meaningfully higher bar for an attacker to clear.

It generally requires either a supply-chain compromise (a device intercepted and tampered with before it reaches the buyer), or a user connecting a fresh, unconfigured device to a computer that’s already infected with malware designed specifically to target BitBox’s USB communication protocol.

For the Silent Payments issue, an attacker would similarly need control of the host machine at the moment a transaction was constructed. In both cases, the physical secure element and offline seed storage that make hardware wallets valuable in the first place were never the point of failure — the vulnerability lived in the software layer connecting the device to a potentially hostile computer.

Tips For Protecting Your Funds

The immediate, concrete step is simple: update your firmware only through the official BitBoxApp, downloaded directly from BitBox’s official site, and never from a link sent via email, social media, or a chat message. Beyond that immediate patch, a few broader practices matter regardless of which hardware wallet you use:

Only initialise new devices on a computer you trust, perhaps on a clean install or a new virtual machine or, at the very least, a new user profile. Since the memory corruption flaw specifically targeted unconfigured devices talking to a malicious host, setting up a brand-new hardware wallet on a clean, malware-free machine closes off that entire attack path.

Buy directly from the manufacturer or an authorised reseller, never from third-party marketplaces or private sellers, and check tamper-evident packaging on arrival. Devices intercepted and modified before delivery remain one of the more realistic ways an attacker gets ahead of you.

Keep firmware current, but verify the update source every time. Firmware update adoption across the hardware wallet industry is notoriously uneven — plenty of people set a device up once, store it away, and never think about updates again. A dormant device running years-old firmware is exactly the kind of target that accumulates unpatched risk silently.

Consider the Bitcoin-only edition if you don’t need altcoin support. BitBox’s own disclosure is a clean, real-world illustration of the “less code, less attack surface” principle — the Bitcoin-only firmware simply didn’t contain the vulnerable memory corruption code at all.

For larger holdings, move toward multisig across multiple vendors, so a flaw in any single manufacturer’s firmware or hardware doesn’t represent a single point of failure for your entire balance.

The Broader Scrutiny Wave Since Coldcard

BitBox’s disclosure doesn’t exist in isolation. Since Coldcard’s RNG flaw became public — ultimately linked to over $112 million in losses across more than 8,600 addresses, traced back to a firmware change from March 2021 that sat undetected for more than five years — the entire hardware wallet industry has been under a magnifying glass.

Trezor and SafePal both disclosed separate data breaches exposing customer shipping and order information belonging to more than 53,000 customers combined. Researchers at Ledger’s own security lab, Ledger Donjon, disclosed a laser fault injection attack against the TROPIC01 secure element used in Trezor’s Safe 7 devices during laboratory testing.

Tangem, for its part, pushed back on the severity of a separately reported flaw, noting the attack required physical possession of the device, invasive lab preparation, and roughly $250,000 in specialised equipment — calling everyday risk to ordinary users “virtually non-existent.”

This is, in one sense, exactly what should be happening.

Increased scrutiny after a major incident is healthy — it’s the security research community doing its job, and manufacturers proactively disclosing flaws before they’re exploited (as BitBox did) is precisely the behaviour users should want to see rewarded rather than punished.

Vendors auditing their own code more aggressively, sometimes with AI-assisted review tools, and publishing detailed technical disclosures rather than staying quiet, is a sign of a maturing industry responding correctly to a wake-up call.

Why This Climate Is a Perfect Storm for Phishing?

But there’s another side to this wave of disclosures, and hodlers are not out of the woods yet. Given the news making the rounds, it’s fair to say that some people will be freaked out and looking for answers, making them a prime target for a constant stream of “urgent hardware wallet vulnerability” headlines; that’s exactly the environment phishing scammers dream of.

Scammers thrive on urgency and fear.

A user who has read three or four headlines in a month about hardware wallets being hacked, seed phrases at risk, and six-figure or seven-figure thefts is primed to react quickly and anxiously to anything that looks like an official security alert — which is precisely the psychological state attackers want you in before they send a fake “critical update” email, a spoofed firmware download link, or a message impersonating a wallet manufacturer’s support team.

The mechanics of this danger show up directly in the guidance every legitimate vendor has had to repeat during this period:

BitBox explicitly warned there is “no question here of downloading a firmware file sent by email, messaging, or social network,” and its own blog post following the COLDCARDs disclosure specifically told users to “stay alert for phishing attempts that use security news to create urgency.”

That warning exists because it’s already happening — attackers watch for genuine security disclosures and immediately spin up convincing fake alerts riding on the same news cycle, counting on panicked users to skip their usual caution.

The pattern to watch for is consistent: an unsolicited message referencing a real, recent vulnerability; a sense of urgency (“act now or risk losing your funds”); a link to download a firmware update, enter your seed phrase for “verification,” or connect your wallet to a third-party site.

Every single one of those requests is a red flag.

No legitimate hardware wallet manufacturer will ever ask for your seed phrase, and firmware updates should only ever be pulled through the official companion app or the manufacturer’s own verified website — never a link sent to you.

Disclaimer: This article should not be taken as, and is not intended to provide any investment advice. It is for educational and entertainment purposes only. As of the time posting, the writers may or may not have holdings in some of the coins or tokens they cover. Please conduct your own thorough research before investing in any cryptocurrency, as all investments contain risk. All opinions expressed in these articles are my own and are in no way a reflection of the opinions of The Bitcoin Manual

Leave a Reply

Related articles

You may also be interested in

Bitcoin Purity

What Is Bitcoin Purity?

Bitcoin Purity is the latest, and arguably strangest, offshoot of the BIP 110 saga. Where the BLAKE2b hard fork responded to BIP 110’s failed soft

What Is Replay Protection?

What Is Replay Protection?

When a blockchain network like Bitcoin splits into two chains through a hard fork, the two incompatible networks face a critical engineering question that determines

Cookie policy
We use our own and third party cookies to allow us to understand how the site is used and to support our marketing campaigns.