Bitcoin Bots Reacted in Seconds to a Deliberately Exposed Wallet Flaw
Researchers at Belgium’s University of Liège deliberately reproduced a legacy Bitcoin wallet mistake. Automated programs reacted within seconds, while the vulnerable addresses once held balances totaling about $106 million at November 2024 prices.
CryptoMEV Bot Newsroom3 min read

Key takeaways
- Two University of Liège researchers deliberately reproduced a legacy wallet-software mistake.
- Automated programs reacted within seconds.
- Vulnerable addresses once held about $106 million, but the actual amount stolen was not established.
Researchers at Belgium’s University of Liège deliberately repeated a very old Bitcoin wallet-software mistake to see whether automated programs would detect and exploit it. The response came within seconds, according to the report.
"The answer arrived within seconds." — University of Liège researchers' experiment, as reported by Crypto Briefing
What researchers tested
The experiment placed exposed funds on the Bitcoin network using a known wallet-software error. Automated programs moved toward the funds almost immediately. The researchers linked the flaw to a group of vulnerable addresses that were scanned by bots.
The bots are believed to monitor the network continuously. They appear to identify wallets that make this particular mistake and attempt to withdraw funds before the owner notices. The timing is the central operational finding: the response was measured in seconds rather than hours or days.
This was not presented as a new vulnerability in the Bitcoin protocol. The issue was traced to legacy mistakes in wallet software, meaning the danger was associated with the tools used to manage coins. The report also notes that old wallet implementations can remain in circulation after newer versions are available, including when users do not update or projects stop maintaining their software.
| Figure | Value |
|---|---|
| Researchers involved | 2 |
| Peak combined balance of vulnerable addresses | About $106 million |
| Valuation date for that balance | November 2024 prices |
| Bot response to the deliberate error | Within seconds |
Exposure, not a loss total
The roughly $106 million figure describes the highest combined balances held by the vulnerable addresses, valued at November 2024 prices (TradingView). It is not a tally of losses, and the researchers did not establish how much was actually stolen from those addresses.
The number should therefore not be read as the amount recovered by attackers or the value lost by any single wallet owner. It represents potential exposure across many addresses and over time rather than a single breach.
That distinction matters for automated traders who examine on-chain movements. A large peak balance can indicate why a security pattern attracts persistent monitoring, but the available report does not provide a basis for calculating realized gains. The response in seconds is the more concrete observation.
Why old errors persist
The findings show that documenting a software mistake does not necessarily remove it from the transaction-monitoring environment. The report says the researchers found bots responding to a deliberately recreated legacy flaw, suggesting that automated scanners still encounter and test addresses associated with that behavior.
For readers following automated Bitcoin strategies, the event connects wallet hygiene with the behavior of on-chain bots. An automated Bitcoin arbitrage bot depends on scanning activity around the chain, while the University of Liège experiment shows how quickly a known address pattern can attract automated responses. The finding concerns wallet software and scanning behavior, not the Bitcoin protocol itself.
The same distinction applies when comparing strategies described in a guide to building an MEV bot. Detection speed alone does not establish profitability, and the approximately $106 million figure does not quantify what bots earned. The report leaves the actual amount stolen undetermined.
Implications for traders
The direct impact is on risk assessment. Large balances tied to a known software behavior can remain visible to automated programs even after the underlying mistake has been documented for years. The speed of the response shows that operators and wallet teams face scanning that operates on a seconds-long scale.
For traders, the event is a warning to separate three figures: peak exposure, confirmed losses and bot revenue. The sources support only the first as a measured amount. They do not establish the latter two. They also do not report a named bot operator, a specific profit figure, or a new protocol-level exploit.
The researchers’ trap therefore produced an operational signal rather than a verified theft total. Legacy wallet errors remained detectable, and automated programs reacted before the owners of ordinary vulnerable addresses had time to respond.
Sources
- University of Liège researchers bait Bitcoin bots with an old wallet mistake — cryptobriefing.com, 2026-11-20
- University of Liège researchers bait Bitcoin bots with an old wallet mistake — tradingview.com, 2026-11-20



