When the Internet arrived, visionaries called it ‘Athens without slaves’ (Rheingold, 1993), so convinced were they that it would enhance democracy and equal representation by giving everyone a voice and a platform. Two and a half decades later, it is common to blame the Internet for political polarization and the rise of demagogues, amongst other democratic and societal threats, and to call for actions to curb it (Wong, 2021). In trying to understand how such rosy predictions could have gone so awry, society looks to mental health experts. After all, it is the interaction between technology and human psychology that is encouraging certain online behaviours or discouraging others. For example, why does reasoned debate – a democratic necessity – quickly devolve online into inflammatory attacks and counterattacks? Or how do anonymity and invisibility over digital platforms become shortcuts for antisocial outpourings? The field, alas, has few explanations to offer. To no small degree, this is due to the rather myopic focus on ‘internet addiction’ and the closely linked ‘Internet gaming disorder’ that has characterized much of the research into the psychological effects of Internet-related technologies and, more broadly, ‘screen time’. We may be paying the price for this narrow focus today as concerning psychological processes, ostensibly unleashed online, transform our world, whilst we watch without being able to explain, much less contain, them. A brief history of how this focus evolved, as well as of early unheeded red flags about other potential online problems, helps explain our current predicament.

We owe the description of what might be considered the Internet addiction field’s ‘patient zero’ to psychologist Kimberly Young (Young, 1996). In 1996, Young reported on a nontech-savvy 43-year-old homemaker with no prior psychiatric history, who, within three months of going online, was spending 60 hr per week in chat rooms. The patient felt ‘depressed, anxious and irritable whenever she was not in front of her computer’, a state she likened for ‘withdrawal’. Over time, the patient stopped performing everyday chores such as cooking, cleaning and grocery shopping and became estranged from her family and friends. Young was inspired by this and similar cases to develop the ‘Internet Addiction Test’, a widely translated and utilized early screening tool (Young, 1998).

The addiction model has remained the dominant theoretical framework when approaching Internet psychopathology. This has been aided by neuroimaging studies that implicated brain regions known to be involved in substance use disorders (e.g. Ko et al., 2009); genetic polymorphism studies that showed alterations in the dopamine reward system (e.g. Han et al., 2007); and reports of successful outcomes with established substance use disorder treatments such as naltrexone (e.g. Bostwick & Bucci, 2008). The addiction focus would ultimately be ‘codified’ in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) in 2013, where a definition of ‘Internet Gaming Disorder’ was added to the appendix (American Psychiatric Association, 2013). Highly overlapping with the DSM-5 criteria for substance use disorders, the definition included the following criteria: withdrawal symptoms when Internet gameplay ceases; tolerance; unsuccessful attempts to stop or reduce engagement with video games; lying about gaming and continued gaming despite the negative consequences. Similarly, the 11th revision of the International Classification of Diseases (ICD-11) in 2018 introduced ‘Gaming Disorder’, whose criteria also borrowed from addiction definitions (World Health Organization, 2019).

Meanwhile, early warning signals about ‘other’ online psychological problems were being raised and, for the most part, not receiving sufficient attention. For example, Aboujaoude, Koran, Gamel, Large, and Serpe (2006) referred to ‘problematic Internet use’ instead of ‘internet addiction’, highlighting commonalities with impulse control disorders and OCD, and cautioning about the effects on personality, privacy, cognition and democracy (Aboujaoude, 2011). Along these lines, a 2008 study (Buffardi & Campbell, 2008) was amongst the first to highlight the problem of online narcissism. Researchers had 129 college students who used Facebook complete the Narcissistic Personality Inventory, whilst analysing their profile for the number of lines in the ‘About Me’ section, the number of wall posts and the quality of the main picture. Results suggested that users with more narcissistic tendencies were more self-promoting in their picture and had a higher number of interactions with ‘friends’. The authors concluded that ‘the average user will experience a network that over-represents narcissists. … This raises the possibility that … norms of expression on social networking sites will be pulled in the direction of greater self-promotion’ (p. 1309).

Early research also suggested that the Internet, partially by facilitating access, was encouraging impulsive behaviour. This was reflected in its effects on impulse control disorders, including compulsive buying disorder (CBD) and gambling disorder. In a 2009 study of 314 adults recruited from an online clothing retailer’s database (Kukar-Kinney, Ridgway, & Monroe, 2009), 17.7% met the criteria for CBD. Rates from studies done before the online retail age, or from studies that did not focus exclusively on online purchases, showed much lower prevalence of CBD, with the largest study estimating it at 5.8% (Koran, Faber, Aboujaoude, Large, & Serpe, 2006).

Early research into gambling disorders also raised alarms about the potential for the Internet to encourage impulsive behaviour. In a 2007 study in the United Kingdom (Wardle et al., 2007), 9,003 respondents – which included late adolescents aged 16 years or older – completed a postal survey. Sixty-eight per cent of all respondents had gambled in the preceding 12 months, with 6% reporting at least some online gambling. Amongst those who had gambled online, 5% met the criteria for pathological gambling, compared to 0.6% of all respondents.

Whether buying or gambling impulsively, correlation does not, of course, prove causation; it is possible that more impulsive individuals gravitate online, and that the Internet in itself does not make users more impulsive. Still, the correlation is striking, and society would have benefited from a more detailed exploration over the subsequent decade and a half of whether interacting with online media can render some users more urge-driven. From impatient texting and ‘sexting’ using smartphones to ‘hooking-up’ on Tinder to ill-afforded purchases on Amazon, online impulsivity is an experience that many can relate to but that lacks the body of research required to understand its underpinnings or how to control it. The fact that children and adolescents may be especially vulnerable for developmental reasons to impulsive behaviour has been exploited by game-makers who have directed in-game purchases at them, making those, not the cost of games, their main income source (Zendle, Meyer, & Over, 2019).

Aggression may be considered in the ‘DNA’ of online and social media life and has been blamed for an increase in bullying, social divisions, political instability, the rise in extremist ideologies and suicide (Aboujaoude, 2011). Early warnings were raised about the longitudinal effects of violent online video games, although with inconsistent results across studies. A 2008 study (Anderson et al., 2008), for example, tracked over 15,000 students aged 9–18 years in the United States and Japan, assessing them twice during the school year to test whether habitual exposure to violent video games led to an increase in offline violence. Results showed that habitual players early in the school year were twice as likely to be aggressive offline later in the year, a finding that held when corrected for differences in gender and baseline aggression, and that was of similar magnitude in two cultures with very different relationships to violence. With the exception of violent video games and cyberbullying, only limited research has been dedicated to the problem of online aggression, despite the increasing of societal recognition of its pervasiveness and perniciousness.

A PubMed Title/Abstract search for ‘Internet’ and ‘addiction’ yields 2,657 entries compared with only 278 when the same search is conducted for ‘Internet’ and ‘impulsivity’. Similar disparities can be seen when comparing publications on Internet addiction with those on online violence and other topics that are highly relevant to psychological health, including the effects of living in a postprivacy world (Aboujaoude, 2019) and the attentional consequences (Firth et al., 2019) of the fast-paced online life. The addiction paradigm has diverted attention from areas that appear more relevant to society at large: even when their experiences cannot be described as ‘addiction’, many people relate to being more impulsive, more inattentive, angrier, overly concerned with their number of ‘friends’, ‘followers’ and ‘likes’ in a perhaps narcissistic way, or worried about the leak of personal information online. How to manage, channel and control these processes has become a civilizational imperative. This is all the more crucial given that children and adolescents – ’digital natives’ who do not know life before Google – may be more impacted by some of these issues, yet they have no other framework by which to judge online experience.

The cost of narrowly focusing on addiction is three-fold. First, as we have seen, problems such as aggression, impulsivity and narcissism that have been linked to the Internet and have been blamed for negative socio-cultural transformations have received comparatively little scientific attention. Second, individuals who do not show addiction symptoms such as tolerance and withdrawal may feel deceptively immune to the psychological impact of Internet-related technologies. (They may also seem deceptively healthy to the therapists trying to help them.) Third, an underestimation of the breadth of the potential psychological harm delays serious public health efforts to raise awareness and consider interventions, and greatly complicates the global coordination and legislative effort required to face the very powerful purveyors of some of these technologies.

Some reviews have pointed to a relatively limited level of psychological harm, highlighting the discrepancy in findings across research (e.g. Piteo & Ward, 2020; Riehm et al., 2019), the limited quality of many studies (e.g. Piteo & Ward, 2020; Riehm et al., 2019) and the role of mediating factors (e.g. Viner et al., 2019). This raises the prospect that, at best, warnings about Internet-related psychological harm are premature or, at worst, amount to fearmongering (Odgers & Jensen, 2020). However, given the rapidly evolving, ever-more sophisticated nature of many Internet-related platforms, even modest signals of possible damage cannot be ignored. Also, much of the research into negative effects focussed on addiction-inspired metrics such as total time spent engaging with the technologies and associated ‘comorbidities’ like depression or anxiety. To be able to see the forest for the trees, research needs to adequately and timely capture the larger psychological context of digital media use, including issues such as impulsivity, aggression, attention span, reading, writing and privacy.

Existing evidence about the risks, harms and unintended consequences of digital media is limited in its scope considering how broad the psychological impact can be. Reassuring evidence can also become quickly outdated. By failing to take a wider perspective on how Internet-related technologies interface with psychology, and by failing to keep up with the pace of technology, the scientific field is hamstrung in giving advice about potential harms and leading the way regarding the need for safeguards, especially when it comes to children and young people. Whilst citizens (and parents) are more concerned than ever about the range of potential online harms and are demanding accountability from ‘Big Tech’, and whilst politicians say they want to be responsive, a narrow-focussed and stagnant body of scientific evidence about adverse effects can hinder regulatory actions. Fearmongering in the absence of solid evidence of digital harm is unhelpful, yet prematurely reassuring children, adolescents, parents and adults can also be dangerous, as it infers that the adverse effects of digital media have been studied exhaustively and we can now rest on our laurels.

Acknowledgements

L.G. is a Joint Editor of CAMH. The authors have declared that they have no competing or potential conflicts of interest.

Ethical information

No ethical approval was required for this article.