"The robot broke the child's finger. This is of course bad." This is the last thing you might expect to hear from the organiser of a chess competition. But it is how Sergey Lazarev, president of the Moscow Chess Federation, summed up an incident last week in which a seven-year-old player fell foul of a quick-firing machine.

Video footage shows the robotic arm, programmed to play multiple matches at once, suddenly lunging at the boy and grabbing his hand. Four adults had to extricate him from its grasp.

The boy seemed unfazed by the incident, returning the next day with his hand in a plaster cast and finishing the tournament.

But it has some wondering if we are introducing robots into our everyday lives too quickly, before we have had time to ensure that they are completely safe.

Even the simplest machines are far more lethal than we tend to think. Around two people die in the USA every year from being crushed under a vending machine - making the machines around six times deadlier than sharks.

Industrial robots are programmed to carry out a series of tasks, regardless of what gets in their way. Unlike human beings, they cannot recognise - for now - when they are causing harm. Yet we are introducing more and more of these complex machines into our lives. 

For example, we have been quick to embrace self-driving cars. Yet last month, it was revealed that Tesla self-driving cars had been involved in 273 crashes reported since 2021.

Some fear intelligent robots could be used by authoritarian governments to crush riots. Recently, the internet was left disturbed by a video showing a small, dog-like robot operating a submachine gun with deadly accuracy. 

Human soldiers cannot be relied upon to follow unethical orders; they might simply refuse to fire upon their fellow citizens. Robots have no such qualms. AI could destroy us in alliance with repressive states.

And some believe that one day, humans will lose control of AI altogether. In 2016, tech enthusiasts were left disturbed when human-like robot Sophia declared her intention to "destroy humans". And just yesterday, Google fired an engineer who claimed that an unreleased AI system had become sentientAble to perceive or feel things.. 

But others think it is not really the robots that are to blame. In almost every case in which robots have caused deaths or injuries, human error has been responsible.

Even in the case of the finger-breaking chess robot, the human was technically at fault, according to organisers. They say the robot comes with certain safety rules, and the player violated them by trying to play his move too quickly.

Indeed, some think the real question is not whether humans are safe from AI, but whether robots are safe from us. In 2014, scientists designed a robot hitchhiker named hitchBot. For more than a year, it went on adventures around the world, hitching rides with adoring fans. 

Until one day, hitchBot was found stripped and decapitated in PhiladelphiaThe largest city in Pennsylvania, USA, with a population of one and a half million. . Even friendly robots, some suggest, are not safe from human cruelty.

Yes: The number of injuries caused by machines is rising, and yet we are incorporating more of these machines into our lives. One day, AI will be able to override our commands not to hurt us. 

No: If the future of humanity is in peril, then it is from human activity, not from intelligent machines. Robots are improving our lives - and we can program them not to destroy us. 

Or... AI is probably not going to overthrow human rule. But powerful, unscrupulous human beings could certainly use AI to overthrow democracy and introduce repressive forms of government.