A notable incident occurred on the Bitcoin network, which once again reminded us of the hidden risks of automating cryptocurrency operations. One user, running a script to speed up a transaction, ended up losing the entire input amount, paying miners a fee of 1.6 BTC. At the time the transaction was confirmed in block 962,142 (August 12), this amounted to approximately $102,778.
Analysis of blockchain data shows that 160,343,885 satoshis were received as the transaction input. All of these funds went to miners as a reward, while the transaction output remained empty. Naturally, the recipient did not receive a single satoshi.
Root of the problem: unbridled Replace-by-Fee
At the heart of what happened is the Replace-by-Fee (RBF) feature, which allows an unconfirmed transaction to be replaced with a new one carrying a higher fee. This is a legitimate tool for speeding up payments during periods of network congestion, when standard fees can spike sharply within minutes.
However, this time the automated script malfunctioned. Instead of stopping at a reasonable level, the bot created a new version of the transaction every second with an increasing fee. With each step, it "consumed" more and more funds from the balance, until the fee reached the entire input amount.
The Bitcoin protocol does not limit the size of the fee that a sender is willing to pay miners, so the network accepted such a transaction. The block mined by the SpiderPool pool brought its participants nearly 1.82 BTC, and 88% of that amount was a "gift" from a single erroneous payment. For miners, whose profits have been declining for several months in a row, such an event is a real stroke of luck.
Lessons for users: how not to lose funds
This incident is a serious wake-up call for everyone who uses automation in their wallets. I recommend following three simple but critically important rules:
First, disable automatic fee increases via RBF. Manually confirming each increase will force you to check the numbers before sending to the network. Second, set a hard fee cap. Most modern wallets allow you to set a limit in satoshis per byte, and that is enough to stop an uncontrolled script. Third, always conduct a test transaction for a small amount after every software update.
The situation also raises an important question for wallet developers: should the interface even allow a scenario where the fee can exceed the transfer amount? Obviously not. The security of automation requires the same attention as protecting your seed phrase—a thoughtless script can drain an account faster than any hacker.
My view: this case is not just a curiosity, but a marker of a systemic problem. Until wallets implement "safeguards" against such scenarios, users will pay for code errors. The market needs smarter solutions in the field of user transaction security.