social media sign

Spinning a Web: Trust and Autonomy on Social Media

read time - icon

0 min read

Dec 15, 2020

The classic character Svengali, from the novel Trilby by George du Maurier, is often referenced as the quintessentially manipulative figure. Having entered the cultural vocabulary to describe a devious, dishonest person, capable of extraordinary manipulation, a Svengali is seen as crafty and clever, and fundamentally untrustworthy.

In the novel, Svengali is a hypnotist, who seduces and exploits the helpless, innocent title character, Trilby.1 Preying on her vulnerability, Svengali demonstrates no conscientious apprehension in taking advantage of her, and turning her into a great, but entirely dependent, singer. An illustration by du Maurier, released shortly after the novel’s publication, portrays Svengali as a spider, spinning an intricate web, used to metaphorically demonstrate his cunning guile and talent for entrapment.2

The relationship between the fictional characters continues to be compelling as a story of exploitation. Trilby is a tragic figure not because of her difficulties, but because she is ultimately not the author of her own story. Svengali’s intervention in her life strips her of her agency, making her prey to his wishes and desires.

While du Maurier uses a literal web to depict this manipulation, various experiences on the virtual web are now raising similar concerns. The recent documentary, The Social Dilemmaportrays a community of Svengali-like Silicon Valley software engineers and social media platform designers.3 According to the documentary, influential strategists from companies like Facebook and Instagram have carefully manipulated their users’ experiences to elicit chemical reactions in the brain that foster dependent behavior from them.

This rapidly emerging dynamic coincides with a growing societal difficulty: while social media usage continues to rise globally,4 so do rates of anxiety, frequently associated with these platforms. This tension, where repeat behavior persists in the face of mounting consequences,5 raises questions about the role of addiction in social media usage.

This article intends to explore that dynamic further. Beginning by outlining how social media platforms are designed to elicit dependent behavior, it will then examine how dependency affects personal well-being, before concluding with a suggestion of how social media designers can potentially restore a sense of autonomy and satisfaction in their users.

Behavioral Science, Democratized

We make 35,000 decisions each day, often in environments that aren’t conducive to making sound choices. 

At TDL, we work with organizations in the public and private sectors—from new startups, to governments, to established players like the Gates Foundation—to debias decision-making and create better outcomes for everyone.

More about our services

Behind the curtain: Manipulation on social media

Our neural circuitry largely responds to rewards and punishments. Put simply, we are motivated to seek out particular experiences or stimuli in our environments that we know can offer us rewards. As we come to learn that a particular stimulus has this property, and this association is reinforced through repeated experience, we become increasingly motivated to seek out that stimulus.6 

To provide a simplified understanding of how this process works, imagine eating your favorite food. As you take a bite, deep within your brain, a chemical called dopamine is released, driving increased cravings and sensations of wanting. Dopamine marks the experience of eating this food as pleasurable, and links it with your surrounding environmental conditions.

As the driver of pleasure, dopamine serves as the chemical incentive to repeat an experience that brought us pleasure initially. That means, if dopamine release is triggered while eating a donut with your morning coffee, the following day, when you sit down to drink coffee again, you will likely crave a donut and feel the urge to replicate that experience.

While that offers a very simplified view of how the brain responds to incentives, the same broad effect can happen with virtual experiences as well. Research suggests that, in extreme cases, adolescents who use social media may experience the six core components of addiction (salience, conflict, mood modification, tolerance, withdrawal, and relapse).7 The platforms’ addictive qualities are largely founded on their unpredictable reward schedules, in which potentially infrequent rewards, largely based on social validation, are doled out via their core functions. 

It’s long been understood in psychology that rewards are much more powerful when they are given intermittently, rather than consistently. For example, when training a dog to sit on command, giving the animal a treat after every round of sitting is much less effective than giving it a treat at random, unpredictable intervals. Similarly, when a user posts a photo on a social media platform,  they cannot be assured of when other users will engage with their photo and offer the social validation that causes a surge of pleasure. Critically, this all operates beyond the user’s control, as the surge of dopamine is tied to the social behaviors of others. As a result, users develop addictive tendencies, compensating for their lack of control with higher rates of engagement that facilitate greater opportunities for social validation.   

Facebook’s “like” button serves as a prime example of a validation function on a social media platform. With this quick access to external validation, within the highly socially-sensitive adolescent cohort, teenagers are given small jolts of dopamine as they receive likes or shares, or any other function of social relevance and acceptance. Interestingly, studies have shown that the anticipation of this dopamine hit, driven by social validation, is potent enough to nearly match the reward itself, in terms of satisfaction.8 

Detractors and critics, like the expert commentators in The Social Dilemma, have taken issue with this deliberate design approach. They argue that social media platforms are being designed to exploit users’ desire for social approval and acceptance, particularly amongst highly vulnerable adolescents. From this position of vulnerability, anticipation, and dependence, anxiety in adolescents continues to rise at a meteoric rate.9 

This trend towards increased dependence on social media platforms presents an interesting question: is it possible to truly take satisfaction from the social media experience without being in control? And how can we reconcile this lack of control with the amount of information users are asked to absorb and assess on social media platforms?

Autonomy: The missing ingredient

Imagine riding a bike and struggling to feel as though you were in control of your direction. Every time you turn the handlebars, the bike resists, refusing to cooperate. Despite having your sights set on a particular destination, the bike seemingly maintains a mind of its own, leading you someplace else entirely. That situation alone is unnerving; it places you in a position of vulnerability to the machine. Should the bike spontaneously swerve off the road or plunge headfirst into oncoming traffic, you would be exposed to the consequent harms. 

There’s no guaranteeing that a collision would be the inevitable outcome. Equally, however, there’s no ensuring that you would be able to prevent that from happening. While that sense of helplessness is enough to provoke anxiety in most, imagine the bike also has means of communicating with you, implicitly. While you initially resist and battle with it for control, it slowly begins to offer you feedback that its course of direction is appropriate. Your instincts begin to waver; your wants and desires are thrown into a state of flux. These messages from the bike are repeated time and again, sending you spiralling into a state of apprehension, uncertainty, and confusion, until the bike has convinced you that its way is best.

This scenario isn’t portrayed in an attempt to stoke fear, but rather to heighten awareness and empathic understanding of the vulnerability, anxiety, and dissatisfaction that comes with a lack of control. As the previous sections outlined, dependency and addictive tendencies resulting from social media usage are rising. Research has shown us that dependency, on a wide range of substances and behaviors, can cause a disruption in core cognitive functions. Attentional bias, for example, suggests that individuals exhibiting dependent tendencies will pay disproportionate amounts of attention to the stimulus triggering the behavioral dependence.10 This has implications on their preference generation, on their decision-making, and on their subsequent behaviors and actions.

All of this is sure to have a detrimental effect on people’s overall well-being. Self-determination theory, presented by the psychologists Richard Ryan and Edward Deci, points to three complementary qualities that ultimately underlie an individual’s well-being: mastery, relatedness, and autonomy.11 In examining the seemingly bleak picture painted above, it is immediately apparent that autonomy—the perception of self-directed freedom, or control over one’s own actions—is grossly lacking. Actions on social media are overtly primed, sequences of movement are scripted, preferences are predetermined. As users, the feeling that we are not acting according to our own volition is unsettling, and contributes to the stress and anxiety many of us now feel with regards to social media.

Who’s in charge?

Recent developments in consent procedures provide a ray of sunshine in improving user engagement with social media platforms. Dynamic consent, a procedure dedicated to greater user involvement and control over their engagement in scientific experiments,12 offers an alternative vision to manipulation and vulnerability. By providing participants with greater decision-making authority with respect to how their personal data is used and distributed, dynamic consent attempts to re-instill a sense of awareness, control, and autonomy in participants’ experiences.

Conventional consent models have traditionally operated on either a single, blanket consent format, or according to an opt-out model.13 While these models technically involve the user in agreeing to the procedure, they often fail to offer the nuance and flexibility that allows the user to feel a sense of control and autonomy over exactly how their information is being used. For example, participants are not given precise authority on where their data is being distributed once it’s been collected. Rather, participants offer overarching consent that ultimately grants the experimenter control over how the data is disseminated. 

Dynamic consent rectifies some of these decision-making imbalances, involving participants in a continuous consent procedure, where their consent is requested at several stages throughout the experimental process. In this, participants are given more control over their data, and generally experience greater autonomy in their experiment engagement.   

Introducing a form of dynamic consent to the user experience on social media platforms may provide a solution to this problem. Adopting the dynamic consent procedure, where users continuously select how they would like to engage with the functionality, data collection practices, and advertising exposure of a given social media platform, offers a potential solution to the lingering problem of the lack of trust and autonomy in the user experience. Choosing to disable the like button, for example, should users feel uncomfortable with its effects, would allow the individual to exhibit the autonomy that ultimately underlies satisfaction. In this, users are once again allowed to take control of their experience and determine precisely how they would like to engage with the platform, including opting out of the features that potentially facilitate addictive tendencies.

Broadly speaking, this approach offers an alternative vision of the user experience. If the traditional social media experience is defined by user vulnerability, the hope is that introducing dynamic consent will prompt feelings of balance, control, trust, and satisfaction in users.


  1. Trilby: novel by du Maurier [Internet]. Britannica. [cited 2020 Nov 5]. Available from:
  2. Svengali as a spider, in his web [Internet]. 1931 [cited 2020 Nov 5]. Available from:
  3. The Social Dilemma [Internet]. [cited 2020 Nov 5]. Available from:
  4. Number of Social Media Users in 2020: Demographics & Predictions – [Internet]. [cited 2020 Oct 25]. Available from:
  5. Addiction | Psychology Today Canada [Internet]. Psychology Today. [cited 2020 Nov 5]. Available from:
  6. Robinson TE, Berridge KC. The Neural Basis of Drug Craving: An Incentive-Sensitization Theory of Addiction. Brain Res Rev [Internet]. 1993 [cited 2018 Jun 8];18:247–91. Available from:
  7. Griffiths MD. Adolescent Social Networking: How Do Social Media Operators Facilitate Habitual Use? Educ Heal. 2018;36(3):66–9.
  8. Morgans J. Your Addiction to Social Media Is No Accident [Internet]. Vice. 2017 [cited 2020 Nov 5]. Available from:
  9. McCarthy C. Anxiety in Teens is Rising: What’s Going On? [Internet]. American Academy of Pediatrics. 2019 [cited 2020 Nov 5]. Available from:
  10. Williams JMG, Watts FN, MacLeod C, Mathews A. Cognitive Psychology and Emotional Disorders [Internet]. Oxford: John Wiley & Sons Ltd.; 1988 [cited 2018 May 11]. Available from:
  11. Ryan RM, Deci EL. The Darker and Brighter Sides of Human Existence: Basic Psychological Needs as a Unifying Concept. Psychol Inq [Internet]. 2000 [cited 2020 Oct 8];11(4):319–38. Available from:
  12. Budin-Ljøsne I, Teare HJA, Kaye J, Beck S, Beate Bentzen H, Caenazzo L, et al. Dynamic Consent: a potential solution to some of the challenges of modern biomedical research. BMC Med Ethics [Internet]. 2017 [cited 2018 Apr 18];18(4). Available from: 
  13. Holm S, Ploug T. Big Data and Health Research – The Governance Challenges in a Mixed Data Economy. Bioethical Inq [Internet]. 2017 [cited 2018 Apr 18];14:515–25. Available from:

About the Author

Stephen Cantarutti

Stephen Cantarutti

Stephen Cantarutti is a Staff Writer at The Decision Lab. He is interested in behavioural economics, and has developed expertise in trust-related decision-making. Before contributing to The Decision Lab, Stephen worked as a Behavioural Science Researcher at Ipsos MORI, applying theoretical behavioural research to develop insights into customer behaviour. Stephen is an avid literature enthusiast, spending much of his free time reading and writing fiction. Stephen has a Bachelor's of Business Administration degree from the Schulich School of Business at York University, a Master’s of Science Degree in Behavioural Economics at City, University of London, and is pursuing a PhD in Behavioural Psychology at City, University of London.

Read Next


The “Social Dilemma” Dilemma

Netflix's social media docudrama has made "dark tech" a hot topic. How can behavioral scientists balance ethics with persuasive design?

digital art of a hand pointing

Combatting Digital Addiction

The Covid-19 pandemic has led to an emphasis on digital communication methods, and for many, a deepening technological addition.

Notes illustration

Eager to learn about how behavioral science can help your organization?