Bitcoin fork Bitcoin Gold suffers 51% attack; attacker with more than half of network's hash rate double spent BTG coins on exchanges and may have stolen $18M
Josiah Wilmoth / CCN :
Context & Ripple Effects
Bitcoin Gold was born in late 2017 as the second major Bitcoin fork, launched deliberately without initial exchange support and with code held back until November, before mining began on a chain designed to thwart specialized mining chips and keep hashing in commodity GPUs. That design choice left the network's total hash rate shallow enough for a well-funded attacker to rent or acquire a majority.
The May 2018 attack is the first large-scale proof that this weakness is exploitable at scale: the attacker used majority control to reverse deposits on exchanges, converting the fork's anti-ASIC philosophy into an $18M liability for the venues that listed it.
First-order effects
- Exchanges that credited BTG deposits during the reorganization absorb losses of up to $18M in double-spent coins, since the attacker sold coins on one side while reversing them on-chain.
- Bitcoin Gold's core developers face an emergency response — raising confirmation requirements or checkpointing — because the chain itself cannot out-hash the attacker.
Second-order effects
- Bittrex's subsequent delisting of Bitcoin Gold shows exchanges repricing small-cap fork listings as uninsurable counterparty risk rather than free trading volume.
- Other GPU-mined forks with thin hash rates come under the same scrutiny, forcing them to add centralized checkpoints or lease hash power defensively to deter copycat attacks.
Third-order effects
- The episode hardens a structural lesson from earlier exchange failures like the Bitfinex hack: in crypto, the custodial intermediary ends up bearing protocol-level risk, giving exchanges effective veto power over which proof-of-work chains survive.
- If shallow-hash-rate forks keep proving attackable, the industry drifts toward fewer, deeper-liquidity chains — decentralization measured in hash rate, not in who can run the miner.
The trend: Small proof-of-work forks are discovering that ASIC resistance without deep hash rate is an attack surface, and exchanges are becoming the gatekeepers that decide which chains stay listed.