Recently, there was an incident that demonstrated how critical blockchain architecture choices can be for user security. One person lost nearly $50 million in USDT due to a single click — and here’s the gist.

It all started simply. The user withdrew funds from a major exchange, and $50 million arrived at their address. The wallet had been active for two years, mainly used for USDT transactions. Nothing unusual at first glance. They first sent a test amount — standard practice, right? After about five minutes, they decided to send the entire amount. But they chose the wrong address.

Here’s where the trap was. The scammer had preemptively conducted a poisoning attack on the address — sent a small USDT amount from an address that was deliberately crafted to look like a real address the victim had used before. When the user copied the address from their transaction history, they accidentally selected the poisoned version. The result — a total loss of $50 million with a single click.

Charles Hoskinson, founder of Cardano, immediately commented on the situation, pointing out the fundamental difference in architectures. In his view, such an incident simply could not happen in UTXO-based systems. And he’s entirely right. Hoskinson noted that this is yet another reason why UTXO is a more resilient model.

The thing is, Ethereum and most EVM networks use an account-based model, where addresses are represented in history as simple text strings. Wallets literally encourage copying addresses from previous transactions. Hackers are well aware of this and exploit it.

Bitcoin and Cardano operate differently. They are based on the UTXO model, where each transaction creates new outputs. Wallets construct operations by explicitly selecting UTXOs, not reusing endpoints. There’s no constant address state that can be visually poisoned. Copying addresses from history simply isn’t necessary.

Hoskinson correctly pointed out that this wasn’t a protocol vulnerability or a smart contract bug. It was a design that interacts with human nature. And in less than an hour, someone lost $50 million because of it. When you see incidents like this, it becomes clearer why architectural decisions at the fundamental blockchain model level matter. It’s not just a technical debate among developers — it’s a matter of the real security of users’ funds.