Where does responsibility lie for deaths caused by viral online challenges?

After a collision that took place on Wednesday night, a five-year-old boy has died, while his mother and three-year-old sister are in critical condition. Their car was struck by a vehicle that, according to investigators, was carrying five young people who were participating in a viral online challenge. This particular social media challenge consisted of driving uninterruptedly for a period of 50 hours.

The little boy and his family are the latest victims of absurd challenges that are widely circulated on social media. This event reopens the debate about the responsibility that these challenges bear for accidental deaths, when there’s a legislative vacuum around the subject and no reliable statistics.

The accident — according to EFE, a Spanish international news agency — occurred on Wednesday night, in the suburb of Casal Palocco in Rome, Italy. The five young people were travelling in a luxury SUV. They collided with a small vehicle, which was occupied by the three members of the family who were the victims of the accident.

The five young people — four males and a female, all in their 20s — used social media to gain popularity, filming themselves while riding around in luxury cars. All indications suggest that, during the accident, they were participating in a viral challenge that consists of driving for 50 hours straight, without taking a break. Their YouTube channel, which promotes various challenges, has accumulated 600,000 subscribers and 152 million views since 2020.

This event reopens the debate about who bears responsibility for the sometimes-deadly consequences of viral challenges that are spread on social media. While there are no reliable statistics, the trickle of tragic cases is constant. Legislation has gaps in how to address this phenomenon, with the laws only able to deal with the physical events that result from the viral challenges.

Gabriel González, who prosecutes cyber crimes in Seville, Spain, has posed the following questions at a meeting on public safety: “What is to be done about viral challenges? What happens when an underage person gets hurt from a [dieting] challenge, or from sticking a condom up their nostril, or seeing who can swallow the most deodorant?”

“We have to see [who should be] penalized,” he warns. “In the real world, it’s not the arms manufacturer who is responsible for a murder, but the one who commits [the crime].”

In these cases, González explains that “in the criminal sphere, what’s not in the legal code cannot be prosecuted. You have to go on the basis of reforms… the latest ones are aimed at introducing new crimes that are not labelled as such, or that don’t perfectly fit into the regulations.”

Legal Vacuum

Cecilia Danesi, a lawyer specializing in liability from the use of artificial intelligence and the author of The Empire of Algorithms, agrees with the prosecutor about the current legislative vacuum. She points out that the European Union’s proposal for AI — within the list of unacceptable risks — will prohibit “systems that are used to manipulate people’s behavior.”

“Imagine that I’m a pervert and I want children to commit suicide, and I put my [human capacity] towards this goal. But imagine that, in addition to this, I use a generative artificial intelligence model to help me carry out [this crime] and that model is then replicated. The viralization of content on social media also involves algorithms, because they’re in charge of deciding absolutely everything we see.”

In this sense, Yieng Xie, a researcher on social media at the University of Texas, states that “these [AI and social media] companies should consider how their recommendation policies stimulate increased activity on their sites. For example, Facebook’s suggestion features are generated by algorithms based on content in which users have previously expressed interest in, as well as on past actions they’ve taken on the platform.”

Like the prosecutor, Danesi explains that, in the criminal area, “illicit conduct has to be 100% consistent. There’s no room for interpretation.” In this way, if the behavior isn’t part of the legal code, it cannot be prosecuted or judged. “In the civil area,” she clarifies, “reparations can be requested.”

In the case of those killed as a result of viral challenges, there’s a question about who is at fault: the person who created the challenge, or the platform that broadcasts it? “The person who creates the challenge is responsible,” Danesi replies, “because their conduct results in harm to another person. The main problem is the causal link. It can be argued that, when creating the challenge, [the person] never thought that it would end in death.” But Danesi explains that the platform also has a responsibility, because it must supervise the content that’s being disseminated. However, with the current systems in place, it’s impossible to monitor everything that is published, especially when, according to Danesi, the challenge doesn’t arouse suspicion at first. When it only becomes dangerous over time, it’s more difficult to detect.

Why do these challenges go viral online?

Francesc Núñez, a sociologist specializing in human emotions and a professor of Humanities at the Open University of Catalonia (UOC), explains in a paper why these challenges are followed and viralized: “Many have a dimension of challenge, whether it’s competing against someone or yourself, or achieving something that requires agility, intelligence or dexterity. To the competitive dimension that the challenge has, we add the personal stimulus of fun and, also, of prowess.”

For Mireia Cabero, a professor of psychology at UOC, there are three main reasons for someone to participate in a viral challenge: solidarity (in the case of challenges meant to raise awareness about a problem), the feeling of belonging to a group, or group entertainment, since many must be done as part of a team… as in the case of the fatal accident in Italy.

Problems arise with dangerous challenges that put the integrity of the participants and third parties at risk. “These [social media companies] are very well built; they work just like a game in stages,” says Silvia Sivera, a professor at the UOC’s Faculty of Information and Communication Sciences. “They naturally choose a younger type of audience — especially adolescents — who share the challenges with their social media networks.”

There are no available statistics about the dangers of these challenges. However, a study carried out by the International University of La Rioja and the University of the Basque Country determined — after surveying 417 minors from three provinces — that one in 10 Spanish adolescents admits to have taken part in dangerous viral challenges.

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition