If you have spent any time at all on the internet, you have encountered internet bots. Web users often don’t fully understand what they are, but they actually outnumber humans online.
The simple definition of an internet bot is a piece of software that runs simple tasks such as reading a webpage and collecting data. But they do so much more. As the internet ages and this technology becomes more sophisticated, bots have been extremely specialized to perform a large variety of tasks and have been embraced by thousands of companies for marketing and customer outreach. But they have also been embraced by hackers.
Bots range in function. They can help you order a pizza on Twitter, book a flight through a messenger app, buy products from eBay and chat with you. “Super bots” such as Apple’s Siri and Amazon’s Alexa are smarter bots people use daily that function essentially as a data collector and communicator. Bots can be programmed to give weather updates at certain locations and times, to sell shoes or be an aid (or even a replacement) for some customer service jobs.
“They can replace help desks to an extent,” said Kade Kerschen, junior in management information systems and IT Help Desk technician. “[Bots] will provide basic ideas like turning off your computer and turning it on again … or can redirect people almost like a phone system.”
Jennifer Tidball, writer and editor for Kansas State’s Division of Communications and Marketing, said there is no researcher at the university who covers this topic.
Bots are used for good and for bad
Bot traffic, according to the Imperva Bot Traffic Report of 2016, made up for 51.8 percent of all internet traffic last year, which is down from 2013 when they made up for 61.5 percent. It has been a consistent trend that about one third of internet traffic is made up of so-called bad bots.
Companies are increasingly using the good ones for long, repetitive tasks and services that bots can do faster and cheaper than humans. But cyber attackers have access to bad ones that act as a tool to broaden the scale of attacks, “casting a wide net with automated attacks” that target thousands of domains at a time, according to the Imperva report.
Good bots are those that generally have official monikers and are designed to be responsible for a task such as curating results in a search engine or optimizing an app for a mobile application. Bad bots are most often impersonators, which masquerade as either a human or a good bot, but they also include the bots that aim to find and exploit vulnerabilities in a website and those that insert spam links into forums.
Impersonator bots, which made up for 24.3 percent of internet traffic in 2016, can sometimes be difficult to spot but the elementary versions, which still account for a large amount of bot activity, are not very tricky.
“There are specific ways they use words that can give them away,” Kerschen said. Some of these bots are given away by their “weird forms of grammar … or an excess of posts that is way more than a regular person would post.”
Tony Ngassi, junior in biology and IT Help Desk technician, said bots on Tinder, the dating service that matches people based on location, are easy to identify.
“Their pictures never have the same person,” Ngassi said.
He said “there will be like five different people” among the preview pictures on the website, and when they are posed next to a landmark far away, it’s clear they are not in the Manhattan area.
Under the umbrella of bad bots are spammers, also sometimes called spam bots or sock puppets, on social media feeds and forums like 4chan that are essentially fake accounts. Some are more malicious than blank accounts that exist to add numbers to another account’s following. Their function is often to spread propaganda, and they do a good job.
Bots spread fake news
A Pew Research study from May 2016 found that six in 10 Americans get news from social media, especially from Reddit, Facebook and Twitter.
The general purpose of lots of bots in social media is to pose as humans and generate clicks and interest to their cause and they dupe people every day. Tweets and Facebook posts from bot accounts linking to bogus stories shape public opinion and affects democracy in the United States and elsewhere.
The presidential elections in the U.S. and France have both been largely influenced by bot behavior, according to a study conducted at the University of Southern California. Using new technology called a Botometer, the researchers collected a massive dataset of 17 million tweets related to the 2017 French election that spread disinformation.
The researchers found that the fraction of social bots in the French MacronLeaks disinformation campaign was 18 percent, which is just 3 percent higher than the results of their study on the disinformation regarding Trump in the Unites States.
The researchers show data that “suggests that the largest majority of [bots] involved in MacronLeaks had prior interests in American politics. … Bots involved in MacronLeaks were also active in the alt-right campaigns leading to the 2016 U.S. Presidential election.”
Twitter has received flak for what many see as the company not adequately addressing the spam bots that plague the site. Perhaps with access to new tools like Botometers it will become much easier to find and eradicate these spam bots from our feeds, but there is currently not a huge conversation about bot removal.
“I think if Facebook or Twitter cracks down, they could be eradicated, but they’re hard to spot,” Kerschen said.
Bots tricking us humans into believing they are legitimate news-tellers is the main way bots can actually do harm to us, though some impersonator bots are also known to hack users’ accounts. It can be difficult to do every time, but we all need to be double-checking the legitimacy of the person (or bot) making posts before we follow links and share them.
