A new report from AI-powered blockchain analytics firm Anchain reveals that bots made up 51 percent of accounts and 75 percent of transactions on the EOS blockchain in the first quarter of 2019 — amounting to a total of $6 million per day in fake transaction data.
By mistaking these inflated transaction numbers for genuine volume, Dapp ranking engines are producing false figures, suggesting that the number of authentic users is much lower than previously thought.
From exchanges to Dapps
Focusing on the Dapp platform with the highest number of users — EOS — the report provides no data on other popular blockchain platforms such as Tron and Ethereum.
Gambling Dapps form the overwhelming majority of EOS apps — accounting for 65% of all EOS Dapp transaction volumes, compared to games at 12 percent, and marketplaces at just 7 percent. This is an improvement from 2018 when a report from dapp.com found that gambling Dapps accounted for an incredible 99 percent of total EOS volume.
As EOS has risen to the top of the rankings along with competitor Tron, Ethereum has fallen, but the original Dapp platform still hosts vastly more Dapps in total — with 1845 Ethereum Dapps registered on StateoftheDapps, compared to 251 EOS Dapps. According to dapp.com however, nearly 600 of these Ethereum Dapps can be classified as "inactive", with no transactions made over the past three months.
Meanwhile, EOS Dapps appear to be bustling with activity as the gambling use case proves popular. But instead of people placing bets, the vast majority of activity is carried out by bots — many of which are profit-motivated and designed to capitalize on dividend payments and loopholes in the betting process.
Good bots and bad bots
The decentralized nature of the blockchain industry, says Anchain, "leaves the door open to bots going undetected for extended periods of time." A characteristic that has made it difficult for developers to resist creating bots for both beneficial and malicious purposes.
According to the report, these bots are typically used by three different parties — the Dapp creators, developers of rival Dapps, and regular users with a technical bent.
Most commonly, the bots serve to "increase liquidity of the tokens and prop up, or grow, asset values," a purpose that can make Dapps more user-friendly, but also inflate volume metrics and artificially push applications up through the ranks of websites like DappRadar. This is similar to the process employed by OKEx and other dubious exchanges who use bots to artificially inflate exchange volume data in order to rise to the top of CoinMarketCap’s exchange rankings.
Elsewhere, profit-motivated developers have designed profit-making bots that take advantage of loopholes to automatically capture the dividends provided by gambling Dapps, and destructive bots designed to stall the progress of competitors.
These malevolent bots use a method similar to Denial-of-Service (DoS) attacks, in which the target Dapp is flooded with transaction requests to overload the system and prevent it from working for real users. Hackers also use what Anchain call BAPT (Blockchain APT hackers), to target smart contract vulnerabilities to release funds in a setup described as "like shooting fish in a barrel."
Amidst these 'bad blockchain bots', a range of 'good blockchain bots' are at work, often issued by the exchange itself to fulfill purposes like testing product quality, and generating automated partners to play with users when no human is available.
Regardless of their purpose, these bots all share 'repetitive behavior patterns' that make them identifiable by Anchain, which employs Artificial Intelligence and machine learning algorithms to detect them. The report suggests these bots will eventually be "dealt with in a compliance context" when regulations are put in place. But until then, the company recommends tackling the problem with a range of initiatives similar to those suggested by Huobi, CoinMarketCap, and others to tackle the inflated exchange volume data problem.
These initiatives include investing in "good bots that help improve product quality and increase liquidity" without artificially inflating volume figures, creating ranking engines that are capable of detecting the presence of bots, and crucially, focusing on "organic human user growth" — a task that might take longer than anticipated.