Failed automation turned into colossal losses for a Bitcoin holder. During a transaction confirmed in block 962,142 on August 12, the user paid a fee of 1.6 BTC — approximately $102,778. The recipient did not see a single satoshi: all funds went to miners, and the transaction output remained empty.
Root of the Problem: Runaway Replace-by-Fee
Chain analysis shows one input of 160,343,885 satoshis, which were entirely consumed by the fee. The culprit was an automated script with the Replace-by-Fee (RBF) function. This option allows replacing an unconfirmed transaction with a new one carrying a higher fee to speed up its passage through the mempool. The mechanism is useful during periods of network congestion, when standard fees skyrocket within minutes.
This time, however, the script malfunctioned. Instead of stopping at a reasonable level, the bot increased the fee every second, "eating" more and more funds from the balance. As a result, the fee consumed the entire input amount, and the transfer was included in a block with a zero output. The mining pool SpiderPool, which mined the block, received about 1.82 BTC in fees, 88% of which came from this erroneous transaction. For pools whose profits have been declining for months, such a "gift" is a real stroke of luck.
How to Protect Yourself from Similar Scenarios
This incident is a stark reminder of the need to control automation. First, you should disable automatic fee increases via RBF and switch to manually confirming each change before sending. Second, set a hard fee limit in satoshis per byte — most modern wallets allow you to set this threshold, which will instantly stop a faulty script.
Third, always conduct a test transfer. A few thousand satoshis will show whether the bot is working correctly and prevent a catastrophe. This is especially relevant after software updates.
The situation also raises a fundamental question: should a wallet allow spending all funds on fees within a single transaction? The protocol does not limit the fee size, and miners accept such payments without question. Until developers implement protective mechanisms, users can only rely on their own vigilance.
My opinion: this case is yet another proof that even "safe" tools require manual oversight. Automation in cryptocurrencies is a double-edged sword: it speeds up processes, but at the slightest error, it can devalue assets faster than any hacker. The industry has long needed to implement standards to protect against such scenarios at the wallet level.