The Trezor “STM32 Entropy” Email Is a Phish Wearing a True Horror Story

Open the email. It comes from an address you've seen before, says "Critical Security Alert: STM32 Entropy Vulnerability" in the subject line, and claims Trezor engineers found a design flaw in the chip inside your wallet — one in four devices could be compromised, and your recovery phrase may not have enough randomness. It closes with a link to "fix" it. Urgent, specific, alarming.

Do not click. That email is the attack, not the warning.

The weaponized truth

The message is fake, but its libel is not invented. Trezor's third-party email provider was breached, and attackers used it to send the phish from the company's legitimate infrastructure in a form that passed the SPF, DKIM, and DMARC checks that normally signal a real sender. Trezor took down the active domain and said the email was not from it, warning recipients not to click any link; no user funds or devices were compromised.

Here is the part that matters for your judgment. The scare is scarier than it should be because the underlying story is true somewhere else. In a live exploit earlier this year, a firmware bug in the Bitcoin wallet maker Coldcard weakened seed randomness from 128 bits to as little as 40 bits, leaving it brute-forceable without ever touching a device. Estimates of the amount swept ran from roughly $70 million to more than $100 million across the waves of the attack. STM32 is the chip family inside many hardware wallets, and entropy is the randomness that generates your seed. The words are accurate. The memory is real. The phishing email just transplants a competitor's genuine catastrophe onto Trezor's brand to make its fake one credible.

That is the mechanism: a phish only works when the bait is true enough to look like a real alert. The attacker who can't break the silicon skips the device and goes for the human holding the seed phrase.

Why it landed in your specific inbox

The scariest part is not the email. It's that the fake reached the right people. This is Trezor's third third-party failure in about a month. In August, partner ShipMonk leaked customer names, phone numbers, and addresses, and in early September Trezor disclosed that an additional 67,000 U.S. customers had been exposed. Customers reported receiving phone calls from strangers who knew their name and that they owned a hardware wallet, plus physical letters with QR codes inside.

Run the sequence in order and it stops being coincidence. Attackers harvested your contact details from a shipping breach, then fired a phishing email from a sender you trust, mentioning the one vulnerability — entropy — that already burned real people this year. The device never had to fail. The leak plus a stolen sender domain did all the work. Even BitBox, a separate wallet maker, warned the same day that its newsletter provider was likely targeted, the pattern pointing at shared email infrastructure across multiple crypto companies rather than a flaw in any single wallet.

What you actually control tonight

A hardware wallet's job is to keep your private keys off the internet. Nothing in this incident changes that math for your Trezor — or for any wallet, since the email is fake and your seed was never at risk. The vulnerability here is not the chip; it's the one layer the wallet can't sit between: you, deciding whether to type twelve words into a web page.

So grade the trade the way you'd grade any setup — name the exit before you consider the entry:

1. Don't click the link. Don't answer the call.Wallet makers never ask for your recovery seed; any message that does is a scam by definition.
2. Only act on channels you open yourself — type trezor.io into the browser, or use the official app. Never trust a link inside an email that surprised you.
3. If you already typed a seed into a linked page, that wallet is burned. Move the funds to a fresh wallet with a new seed immediately. Do not update, do not migrate, do not "reset" — migrate.
4. When gauging any hardware wallet vendor, treat third-party exposure as a selection criterion, not a footnote. The chip is not the attack surface that empties retail wallets in 2026; the vendor's providers and your own reflexes are.

The expiry date

This specific checklist expires the moment the phishing domain is taken down and the breached provider is identified — that is cleanup, not wisdom. What does not expire is the read underneath it: on-chain thefts like the Coldcard one get the headlines, but the losses most owners actually face come from social engineering dressed in a true story and a borrowed sender address.

Before you run any "urgent" security step again, check the same three things: does the sender route back to a channel you opened yourself, is anyone asking for the seed, and does the claimed vulnerability actually exist for your wallet — or only for a competitor's? The answer to the last one is why this phish nearly worked. It's also why you now know it won't.