A week-long 51% attack on the Bitcoin network would cost $6 billion, research reveals

Bitcoin is trusted by governments and financial institutions. Reputable investors call it “digital gold,” and some even claim it’s better than gold. Notwithstanding all these, Bitcoin is still facing various security threats. Lately, a lot has been said about the potential threat from quantum computers. However, Duke University Professor Campbell Harvey revealed another concern, namely the relative cheapness of a 51% attack on the Bitcoin network. 

Duke University Fuqua School of Business Professor Campbell Harvey released a paper dedicated to potential threats to Bitcoin. In an abstract, Harvey likens Bitcoin to gold but outlines that Bitcoin is facing its specific threats: quantum computers and, what’s more important, a possible 51% attack. He also recognizes that Bitcoin has its advantages over gold. For instance, he mentions that “modern alchemy” allows the production of more gold, while Bitcoin’s supply cannot exceed 21,000,000 units.

What’s a 51% attack?

As Bitcoin mining is costly and requires special hardware, miners don’t have an opportunity to mess with the ledger data. Each node “votes” via computing power (hashrate) to validate transactions in new blocks, and the majority of miners are voting for correct data. Miners are motivated to vote for the correct data as they depend on the Bitcoin blockchain’s integrity, which gives it value. 

However, once half of the total hashrate in the system is controlled by a single entity (a person or a group of plotters), it has the power to change the records in the Bitcoin ledger. It will allow bad actors to move other people’s bitcoins, effectively stealing them. 

While some criticize Bitcoin for its low decentralization level, no one in 16 years of Bitcoin’s existence has ever managed to gain control over the Bitcoin blockchain. 

In the early days of Bitcoin, mining was accessible to any PC owner. However, as mining is based on competition where the luckiest miner has to have a higher hashrate level than most rivals, computers and even GPUs and FPGAs quickly became obsolete for mining. In 2013, the first ASICs (devices specialized for Bitcoin mining) hit the market. Soon, Bitcoin mining turned into a multi-million-dollar industry, requiring much investment and facilities filled with humming ASIC devices. In October 2025, Bitcoin mining difficulty reached a new maximum.

It makes hacking Bitcoin via a 51% attack a hard and expensive task. As the mining difficulty is going up, each year the costs of a 51% attack are getting higher.

Campbell Harvey’s findings

While a 51% attack is costly, its price is not unthinkable. Such networks as Bitcoin Gold and Ethereum Classic suffered several 51% attacks after 2017. Each one of them resulted in one million plus worth of crypto stolen in each separate case. In August 2025, Qubic mining pool claimed it got over 50% of the hashing power in the Monero network. 

Professor Harvey calculated the costs and concluded that one week of domination on the Bitcoin blockchain would cost “only” $6 billion:

The research is based on the following metrics:

• Bitcoin’s annual output is 164,363 BTC
• Energy usage is 166.4TWh
• Total cost is $12 trillion
• Total energy cost is $8.4 trillion
• Total cost per unit is $73,000 per 1 BTC unit

Harvey noted that a successful 51% attack on Bitcoin would cause a severe price drop, and hackers could still profit from it and earn much more than $6 billion back. Harvey estimated BTC perpetual futures daily volume at $60 billion and conventional BTC futures daily volume at $10 billion. Harvey believes that opening a short position on these markets before a 51% attack could result in high profits for attackers on top of returning $6 billion. Harvey adds that the motive may not be profit-related.

However, critics of Harvey’s warning argued that setting such a huge mining operation would have taken years, and it wouldn’t go unnoticed. More than that, shorting so much BTC in the conditions of an ongoing 51% attack may be hard, as the exchange will probably flag an operation as market manipulation and won’t let it.

Commenting on Harvey’s research, Matt Prusak, president of American Bitcoin Corp., told Bloomberg: “My attitude is that economic feasibility kills the 51% thesis. I live in the real world, and I am not concerned.”