Elon Musk and Twitter are at war over bots - here's why they matter

by Rina Chandran | @rinachandran | Thomson Reuters Foundation
Wednesday, 13 July 2022 08:56 GMT

An image of Elon Musk is seen on smartphone placed on printed Twitter logos in this picture illustration taken April 28, 2022. REUTERS/Dado Ruvic

Image Caption and Rights Information

Musk has terminated his $44 billion Twitter buyout deal, citing doubts over bot info. But what are bots, and are they all bad?

By Rina Chandran

July 13 (Thomson Reuters Foundation) - When Elon Musk announced he planned to walk away from a $44 billion deal to buy Twitter last week, one of the main reasons was the number of bot accounts on the social media platform.

Musk had earlier said that one of his biggest priorities after acquiring Twitter was to "defeat the spam bots or die trying", and he has contested the company's analysis that fake and bot accounts make up less than 5% of its users.

Twitter said its data on spam and bot accounts is correct, and on Tuesday sued Musk for violating the deal.

But what exactly are bots? And are they all bad?


A bot - short for robot or internet robot - is a computer programme that runs automated tasks over the internet. They are often used to carry out basic or repetitive tasks.


Bots accounted for more than 42% of global website traffic last year, according to cybersecurity firm Imperva Research.

Bots are used for a wide range of functions - from posting sports scores to finding the cheapest prices of items online.

Chatbots are commonly used in customer service. They often use artificial intelligence to answer queries in a chat that mimics a conversation with a human operator.

They are also widely used on social media.


Bots can spread useful information such as updates on COVID-19, the traffic and the weather, while chatbots can answer questions.

Google, Bing, and other search engines use "crawler" bots to find new content posted online, while bots also help scrape the internet for research, find shopping deals like cheap flights, and to identify abusive or violent content.

Twitter has a #GoodBots label to identify some that are useful, including the Thread Reader App that allows users to save Twitter threads, and the Endless Screaming bot that adds wailing comments to any Twitter conversation on demand.  

Bots are also increasingly being used in the criminal justice system, though these can be unreliable and lead to bias, critics have warned.

Even good bots "cannot solve unknown or complex queries, and they lack empathy", said Ritesh Chugh, an associate professor of technology at Australia's CQUniversity.


Spam bots or fake accounts are designed to manipulate or artificially boost activity on social media platforms. They can tweet malicious links, spread fake news and propaganda, and steal user data by masquerading as real accounts.

Unlike many 'good' bots, they are often designed to hide the fact they are automated and appear to be real people.

Twitter said it removes more than 1 million spam accounts every day, which includes both bots and other fake or nuisance accounts. Other social networks also remove large numbers of spam accounts.

Nearly 28% of online traffic last year was made up of bad bots, Imperva's "Bad Bot Report" found.

"Bad bots are now more advanced and evasive than ever, mimicking human behaviour in ways that make them harder to detect and prevent," it said.


Imperva's research shows that the percentage of online activity carried out by bots is rising.

Bots are here to stay, as they help companies cut costs, and reduce response times for customers, said Chugh.

"Bots will continue to mature as their usage becomes more widespread ... artificial intelligence and machine learning will improve the performance of bots," he said.

"However, it isn't just the good bots that will improve and survive, but the bad bots too.”

Online platforms are facing calls to improve systems to disclose data on bots and root out bad actors.

Chugh also said that people should be given a right to know when they are not dealing with a real person, calling for rules requiring companies "to disclose to their customers any interactions with an automated bot."  

Related stories:

As Malaysia tests AI in sentencing, lawyers fear for justice

African software developers use AI to fight inequality

Drones to data: Africans turn to healthtech

(Reporting by Rina Chandran @rinachandran; Editing by Sonia Elks. Please credit the Thomson Reuters Foundation, the charitable arm of Thomson Reuters, that covers the lives of people around the world who struggle to live freely or fairly. Visit https://news.trust.org)