Researchers have warned that software robots, also known as benevolent bots, sometimes have online fights and are more like humans than we think. The bots are designed to improve articles on Wikipedia by editing spelling, creating links, voiding any vandalism and enforcing bans.

The online world has transformed into an ecosystem of bots but it would seem these automated employees are showing poor interaction skills. They are unpredictable and this interaction between them is not by design which is causing some worry amongst inventors. Bots do not have the capacity for creativity, emotions and social skills so it would be expected their relationships with one another are uneventful and predictable.

Although the bots are intended to support the online encyclopaedia they often undo each other’s edits and begin `fights` over content that can continue for years. Unlike humans who use Wikipedia the bot’s interactions occur over long periods of time, where as we simply stop by and leave they remain there. However just like humans, bots may behave differently in new cultural environments. Research has suggested that bots with less intelligence may give rise to complex interactions and this has led to important uncovering’s in Artificial Intelligence Research.

The fact bot’s have longstanding conflicts raise the concern that humans are failing to spot the problems early enough and understanding bot-bot interactions is crucial for managing social online interactions successfully. Cyber-safety needs to be tighter and researchers are looking into reasons for such unpredictable outcomes from their products to avoid any strange behaviour in the future.