PRIVACY RISKS AND HARMS The Common Sense Privacy Program Common Sense is the na on’s leading nonprofit organiza on dedicated to improving the lives of kids and families by providing the trustworthy informa on, educa on, and independent voice they need to thrive in the 21st century.
www.commonsense.org INTRODUCTION
The Common Sense Privacy Risks and Harms report iden fies risks to children and students as they engage online and iden fies ways for parents and educators to choose the products that best protect our youngest consumers from privacy intrusions and manipula on by third par es that could have long-term implica ons.
These decisions by parents and educators on which products to use at home and in the classroom need to be guided by resources backed by research and experts with informed analysis of the risks. Our easy-to-understand privacy evalua ons from Common Sense include an overall score, display er risks, and summarize privacy concerns to guide parents and educators in making informed choices. Informa on and communica on technologies offer tremendous benefits to children, especially the most disadvantaged, but parents and educators need to be able to harness the power of the technology while at the same me limi ng the harms in order to protect children. As parents, educators, and consumers, our main leverage in encouraging companies to make changes in how they collect and use personal informa on from kids is in our purchasing decisions, by us only buying products for kids that protect their privacy and avoiding products that do not.
Privacy has meant many things over me, but in the digital age the stakes are high, and the issue raises important ques ons about what personal informa on is collected from kids by the applica ons and services they use every day, how that informa on is used, and with whom it’s shared and why. The understanding of the implica ons vary, and the choices we make for our children now can have ripple effects for decades to come. Many parents and educators say they are not concerned about the right to privacy and believe there are no real privacy risks or harms because they think their children and students have nothing to hide.
However, the choices kids make (and those their parents make on their behalf) with personal informa on are ul mately choices that define their online iden es and profiles. When it comes to privacy, different people face different kinds of privacy risks and harms. Kids are especially suscep ble to behavioral, social, emo onal, physical, and financial risks that could create lifelong social and emo onal harm. For example, when students take college entrance exams such as the PSAT, the ACT, or Advanced Placement exams, they are o en asked to check off a box if they want to receive informa on from colleges or scholarship organiza ons. That simple act of checking a box to share a kid’s personal informa on and their exam grades with third par es can be given without parental no ce or consent and introduces privacy risks. Organiza ons like the College Board and ACT could use a student’s personal informa on to create and market personal profiles to third par es. A student’s personal informa on could also be combined with their online digital footprint to create detailed profiles that may be used by college admissions offices to determine acceptance based on sta s cal analysis of their data that takes into considera on the student’s sex, race, and behavior on social media accounts.
With this report, we collect the best available informa on about ways consumers can arm themselves with informa on when choosing which technology tools to use. There is no one-size-fits-all solu on for privacy, and so parents and teachers need to educate themselves with resources like ours and those offered by other trusted sources in order to best understand how to minimize the risk of harms to our youngest consumers based on the personal informa on collected from them, who has access to it, and how it is used.
The Common Sense Privacy Program
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS Common Sense is grateful for the generous support and underwri ng that funded this report from the Michael and Susan Dell Founda on, the Bill and Melinda Gates Founda on, and the Chan Zuckerberg Inita ve.
CREDITS
Authors: Girard Kelly Jeff Graham Jill Bronfman Steve Garton Suggested cita on: Kelly, G., Graham, J., Bronfman, J., & Garton, S. (2019). Privacy risks and harms. San Francisco, CA: Common Sense Media
This work is licensed under a Crea ve Commons A ribu on 4.0 Interna onal Public License.
PRIVACY RISKS AND HARMS privacy.commonsense.org TABLE OF CONTENTS
Privacy Risks 1
What are the risks? ...... 1
Who is vulnerable to the risks? ...... 2
Three forms of risk: content, contact, and conduct ...... 2
What digital footprints do kids leave? ...... 5
What type of personal informa on is at risk? ...... 7
Internal informa on ...... 7
External informa on ...... 7
Financial informa on ...... 7
Social informa on ...... 8
Technological informa on ...... 8
Poli cal informa on ...... 8
Privacy Harms 9
What are the harms? ...... 9
Social and emo onal harms ...... 10
Surveillance harms ...... 11
Ideological harms ...... 13
Who poses a risk of harm to kids? ...... 13
First- and third-party companies ...... 13
Organiza ons ...... 14
Governments ...... 14
Bad actors ...... 15
How parents and educators can make a difference ...... 16
Conclusion 17
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS
are not always able to discern the difference between mes- PRIVACY RISKS sages meant to convey truthful, accurate, and unbiased infor- ma on and messages meant to influence and change their What are the risks? behavior (American Psychological Associa on, 2004; Nyst, 2017, pp. 8–9). Ul mately, these messages could distort As applica ons and services collect more and more personal kids’ percep on of reality, which could have nega ve con- and behavioral informa on about children at home and stu- sequences on their mental health and might affect their self- dents in an educa onal se ng, it is impera ve that the pri- control and ra onal decision-making processes. vacy of that informa on be protected from a wide range of poten al risks. To begin with, informa on must be protected “There are also clear gaps in children’s from poten al misuse by third par es and from a data breach. knowledge about risks online, and However, kids also face unique privacy risks when trying to despite rapidly increasing usage among determine which applica ons and services are safer to use children and adolescents, many lack with countless websites and app store products all seeking digital skills and the cri cal ability to their a en on. gauge the safety and credibility of Companies o en offer a rac ve free or low-cost apps with content and rela onships they decep ve in-app purchases that are specifically targeted to- experience online” (Nyst, 2017, p. 9). ward kids. These apps o en have worse privacy prac ces— for the purposes of our evalua on process—because they The following examples illustrate some of the different types use the informa on they collect, or the informa on of the of behavioral, social, emo onal, physical, and financial risks parents, to engage in third-party marke ng, adver sing, and that can occur from the misuse or inadvertent disclosure of tracking technologies that could use the collected personal a child, student, parent, or educator’s personal informa on: informa on to profile and target kids—as well as their par- • Behavioral risks include modified personality changes ents and teachers—with adver sements outside of the prod- such as nega ve a tude, early or increased use of al- uct (Kelly, Graham, & Fitzgerald, 2018). These contextual or cohol, parent-child conflict, loss of appe te or interests, targeted adver sements may also influence kids across their inconsistent preferences or beliefs, a en on deficits, devices and the internet over me. The type of risks that kids sudden changes to characteris cs, extreme opinions, ir- face depends on the type of personal informa on collected ra onal inten ons, abandonment of established habits, and for what purposes that informa on is used to cause in- and verbal cyberbullying or abuse of others (Living- ten onal or uninten onal harm. stone, Mascheroni, & Staksrud, 2015).
“Research shows that children under the • Social risks include modified professional or educa- age of eight are unable to cri cally onal changes such as loss of employment, damage to comprehend televised adver sing reputa on, social or emo onal learning problems, poor messages and are prone to accept learning outcomes at school, disciplinary ac ons, ex- adver ser messages as truthful, accurate pulsion, criminal or civil charges, self-destruc ve rela- and unbiased” (American Psychological onships, running away, and separa on from family or friends (Livingstone et al., 2015). Associa on, 2004). • Emo onal risks include modified mental health changes Collec on of personal informa on for the purposes of ex- such as nega ve body image, substance abuse, sexual er ng influence over kids is an inherently decep ve pro- deviance, anxiety, depression, low self-esteem, addic- cess where kids learn at a young age about the principles on, avoidance, isola on, aggression, and lack of empa- of surveillance capitalism and that their informa on and its thy for others (Grabe, Ward, & Hyde, 2008). meaning has monetary value that can be exploited by oth- • Physical risks include modified decision-making pro- ers and exchanged for power (Christl, 2017; Zuboff, 2016; cesses such as ea ng disorders, acts of verbal or Zuboff, 2015). These adver sing messages are intended to physical violence toward family or friends, sexual as- coerce kids into making choices they would not otherwise sault, reckless endangerment, manslaughter, or self- make. The choices that kids make in response to adver sing- harm such as cu ng or suicide (Graafland, 2018). related messages may not always be in their best interests because adver sing messages are designed to exploit kids’ • Financial risks include modified economic consump- suscep bility and could be destruc ve to a kid’s develop- on changes such as impulsive in-app purchases, unex- ing state of mental health. Adver sing messages also serve pected online purchases or gi s, extor on, credit fraud, to exploit kids’ vulnerability in cri cal thinking because they
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 1 large withdrawals, large purchases on credit, applica- that companies can exploit for their financial gain (Crain & ons for more credit, maxed-out credit limits, loss of Nadler, 2017). income, taking high-interest loans, the from family or For example, research from the American Medical Associa- friends, sales of family or sen mental items, pa erns of on (AMA) has found that most free apps designed to help borrowing to make payments, food and housing insecu- people quit smoking or cope with depression share data with rity, and unhealthy spending habits (Meyer, Adkins, & third par es for adver sing purposes (Huckvale, Torous, & Yuan, 2019, January). Larsen, 2019). Research indicates “behavioral science also suggests your mood, energy levels, and alertness can affect Who is vulnerable to the risks? decision-making and biases in predictable ways. According to leaked internal documents, Facebook claims it can iden- Knowledge gaps of the risks are not surprising. Companies fy its teenage users’ emo onal states to give adver sers the inform people incompletely, inaccurately, or not at all about means to reach those who feel ‘insecure,’ ‘anxious,’ or ‘worth- their data prac ces with ambiguous, misleading, and o en less.’ Presumably the point is to pinpoint the exact moment obfusca ng language in their user interfaces and policies a sales message is most likely to hit home” (Crain & Nadler, such as their privacy policies and terms of service. When as- 2017). The types of risk that vulnerable kids face online can sessing vulnerabili es of kids, it is important to take into ac- be categorized by: the content they are exposed to or share count children’s own a tudes about online risks—which of- with others; the contact they have with trusted or untrusted ten differ considerably from those of adults. Although there individuals; and the conduct they display toward others that is a lack of research on some of the most marginalized com- defines their iden es. muni es and groups, Burton, as cited in Unicef (2017, p. 81), provided exis ng evidence that indicates “children who are Three forms of risk: content, contact, and most vulnerable to online harms include girls, children from conduct poor households, children in communi es with a limited un- derstanding of different forms of sexual abuse and exploita- Researchers typically organize the wide range of privacy risks on of children, children who are out of school, children with encountered online into the following three categories: con- disabili es, children who suffer depression or mental health tent, contact, and conduct risks (Livingstone et al., 2015). problems and children from marginalized groups.” • Content risks: Where a child is exposed to unwel- “Digital technology and interac vity also come and inappropriate content. This can include sex- pose significant risks to children’s safety, ual, pornographic, and violent images; some forms of privacy and well-being, magnifying adver sing; racist, discriminatory, or hate speech mate- rial; and websites advoca ng unhealthy or dangerous threats and harms that many children behaviors, such as self-harm, suicide, and anorexia. already face offline and making already-vulnerable children even more • Contact risks: Where a child par cipates in risky com- vulnerable” (Unicef, 2017, p. 8). munica on, such as with an adult seeking inappropriate contact or solici ng a child for sexual purposes, or with Also, how much children benefit from digital experiences individuals a emp ng to radicalize a child or persuade has much to do with their star ng points and opportuni- him or her to take part in unhealthy or dangerous be- es in life. Research indicates “while those with strong so- haviors. cial and familial rela onships are likely to use the internet to • Conduct risks: Where a child behaves in a way that con- bolster these rela onships—leading to improved well-being— tributes to risky content or contact. This may include children experiencing loneliness, stress, depression or prob- children wri ng or crea ng hateful materials about lems at home, for example, may find that the internet com- other children, inci ng racism, or pos ng or distribu ng pounds some of these exis ng difficul es. Conversely, chil- sexual images, including material they have produced dren who struggle with their social lives offline can some- themselves. mes develop friendships and receive social support online that they are not receiving elsewhere” (McKenna, Green, & While it is rela vely easy to categorize various forms of risk in Gleason, 2002, p. 9). These examples are not limited to kids— general, it is much harder to determine the risk rela onship consider all the ways that marketers might be able to iden fy between a par cular online image or ac vity and an individ- when parents break off a rela onship, experience a signifi- ual child. The following table describes these types of risks cant health concern, start a new job, lose a job, or go through and the resul ng harms (Nyst, 2017, p. 73): other events that create a habit-flexible, vulnerable moment
2 PRIVACY RISKS AND HARMS privacy.commonsense.org Content: child as recipient Contact: child as par cipant Conduct: child as actor (mass produc ons) (adult-ini ated ac vity) (perpetrator or vic m)
Aggressive • self-abuse and self-harm • radicaliza on • cyberbullying • suicidal content • ideological persuasion • stalking • discrimina on • hate speech • harassment • exposure to violent • stalking • violent peer ac vity content
Sexual • unwanted/harmful • sexual harassment • child sexual abuse exposure to • sexual solicita on • produc on of child pornographic content • sexual grooming abuse content • child-product indecent images
Values • racism • self-harm • poten ally harmful • biased or misleading • unwelcome ideological user-generated content informa on persuasion
Commercial exploita on • embedded marke ng • viola on and misuse of • livestreaming of child and adver sing personal data sexual abuse • online gambling • hacking • sexual trafficking and • fraud and the exploita on of children • sexual extor on
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 3
What digital footprints do kids leave? “The act of going online can dismantle the tradi onal protec ons most As children come of age in today’s digital society, the digital socie es try to place around children, footprints they leave behind has been shown to have a mea- exposing them to unacceptable content, surable impact on their lives as adolescents and adults (Mar- unacceptable behaviour and poten ally n, Wang, Pe y, Wang, & Wilkins, 2018). Research shows dangerous contacts with the outside that, over the last few years, there has been “a significant world” (Unicef, 2017, p. 71). increase in Internet usage by 0-8 year olds, partly because children start using digital devices at younger ages. On av- Tradi onal protec ons were established by adults to pro- erage across Organiza on for Economic Co-opera on and tect children in real me, either with live adult supervision Development (‘OECD’) countries with access to digital tech- or structurally, to subs tute for adult supervision. In a pre- nologies, 18% of students in 2015 accessed the internet for internet era, adults held the remotes for live channel chang- the first me before reaching the age of 6, an increase of 3 ing when inappropriate content came onto the television, percentage points since 2012” (Graafland, 2018, p. 10). In which was situated in a centrally located living room. Even addi on, parents can be a poten al source of children’s data when an adult was not physically present, parents could as- misuse. Parents increasingly start building digital footprints sume they set a channel or program and could walk away for their children, even before they’re born (e.g., parents an- from the television and the selected program would be ap- nouncing the mother’s pregnancy on social media pla orms propriate for their kids. Furthermore, me-slot programming and apps tracking the movement and heart rate of the fe- prevented younger children, many of whom were likely to tus). Many hospitals even offer access to third-party profes- be in bed by 9 p.m., from seeing mature content. With each sional photographers to take newborn baby photos with the technological innova on, these tradi onal protec ons broke mother and her baby only mere hours a er birth. Unknown down. First, there were mul ple television screens in private to most mothers is that those in mate photographs may be rooms, then recording devices to me-shi programming, used to market professional photography services to other then internet-accessible television programming available at parents and hospitals for the rest of that child’s life. all hours, on mul ple devices, and to any user. Similar transi- Research has shown that 81% of children under age 2 cur- ons occurred with regard to relying on adult playground su- rently have some kind of digital footprint, with images of pervision to break up physical fights and classroom teachers them posted online (AVG Technologies, 2015). In the U.S. reminding students to keep their eyes on the board. Now bul- that percentage rises to 92%, while for the EU the figure lying can occur online outside of adult supervision or knowl- is 73% (AVG Technologies, 2015). When it comes to digi- edge, and students communicate and learn via a variety of tal skills for kids 2 to 5 years old, “more small children can technologies, not all of which are a communal experience open a web browser (25%) than swim unaided (20%), and under the control of a single classroom teacher. most 6- to 9-year-olds and almost half [of] 6- to 7-year-olds The more me kids spend online, the greater their exposure spend more than two hours a week online” (AVG Technolo- to online opportuni es and risks and also the greater their gies, 2015). Kids reach digital maturity by the me they’re digital footprints. Research indicates these go hand in hand, 11, because that is when they establish their own online as children must encounter and explore online risks in order iden es, or begin to par cipate in their online iden es to learn and develop digital skills. A empts to minimize risks created by their parents and graduate onto mainstream so- can limit children’s online opportuni es, while efforts to max- cial networks like Facebook, Instagram, Snapchat, and Twit- imize opportuni es can also increase digital risks (Livingstone ter (AVG Technologies, 2015). Research has shown that 21% et al., 2011, EU kids online). As kids start to reach 14 to 17 of children have encountered websites containing poten ally years old, their online ac vi es are o en completely unsu- harmful user-generated content such as sites containing hate pervised. Research shows that 40% of 17-year-olds secretly messages, anorexia/bulimia sites, sites promo ng self-harm, access their Facebook accounts without parental knowledge or sites that discuss taking drugs. In separate studies in 2011 when told not to use their devices, which is partly due to and 2018, approximately 9% of children age 11 to 16 ex- adolescents being increasingly worried about the number of perienced some form of personal data misuse (Graafland, likes, favorites, or retweets they get on their posts (AVG Tech- 2018; Livingstone, Haddon, Görzig, & Ólafsson, 2011, Risks nologies, 2015). Adolescents o en delete their recent pho- and safety). The most common misuses reported were that tos or posts that do not have a large number of likes or fa- someone had used their password or pretended to be them vorites in order to support the impression that all of their (7%); personal informa on abuse (4%); and that they had lost content and social interac ons are popular, which further money by being cheated online (1%) (Livingstone, Haddon, maintains their online persona, social reputa on, and status Görzig, & Ólafsson, 2011, EU kids online). among peers. Accordingly, approximately 40% of 13- to 17- year-olds in the United States reported feeling pressured to
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 5 only post popular or fla ering content (Lenhart, 2015). This on social networks that have caused them privacy concerns, is not surprising, as receiving “one-click” feedback (such as while 23% limit their posts and more than half (53%) are more likes or pokes) ac vates the part of the brain that is involved careful about what they post” (AVG Technologies, 2015). in explicit pleasure and addic on (Sherman, Payton, Hernan- A recent survey found that “more than half of employers dez, Greenfield, & Dapre o, 2016). The developing adoles- have said they have not hired a candidate because of con- cent brain is more easily able to create pathways of addic- tent they found on social media, but a similar number said if on that involve several factors and processes, as described they can’t find an applicant online, they are less likely to call in Winters, as cited in Nestler & Malenka (2004), that in- that person in for an interview” (Quraishi, 2019). This leaves clude the neurobiology of addic on, environmental factors, young people who are trying to par cipate in educa onal genes, vulnerabili es, and disorders. Research about adoles- and voca onal opportuni es in a quandary. If they have no cent brain development provides insights into and addi onal online presence, they may not only not be able to communi- clues as to why adolescence might be a par cularly vulnera- cate with friends, but they also may not be able to access on- ble period for developing a substance use disorder or inter- line assignments for college classes or job applica ons avail- net addic on to receiving “one-click” feedback (such as likes able through social media sites. Yet if they have an expansive or pokes) (Casey, Jones, & Hare, 2008; Winters, 2009; Felt online presence, such as a paid influencer, video content cre- & Robb, 2016). ator, or controversial blogger, their online persona may clash However, this behavior can be harmful, as receiving “one- with the job opportunity offered by a company that does not click” feedback is associated with reduced well-being (e.g., want to be associated with such content. Needless to say, so- lower self-esteem, increased anxiety and depressed feelings) cial content that indicates that the young job applicant has among adolescents (Burke & Kraut, 2016; Kross et al., 2013). used drugs or par cipated in any illegal ac vity would be a Digital footprints also allow adver sers to learn which con- disincen ve for an employer. Even deleted posts or posts tent a child does or does not favorite or like, which allows with a limited shelf life may come back to haunt the young adver sers to specifically target messages to kids they know job applicant. are more likely to be influenced and also to spread viral con- As kids reach adulthood and enter the job market at age 18 tent through their peer groups. A Pew Research Center sur- to 25, they are presented with new types of risks. As adults, vey showed that 81% of parents of 13- to 17-year-olds sur- today’s children and adolescents will be subject to a scru ny veyed in the United States reported being concerned about and historical digital footprint record that we cannot begin how much informa on adver sers can learn about their to imagine. For example, “while most of us were, blessedly, child’s online behavior (Madden, Cortesi, Gasser, Lenhart, & able to forget, reimagine or reinvent part of our early lives, Duggan, 2012). loves, jobs, thoughts, deeds, comments and mistakes, today’s children will be in a very different spot. A single stupid com- “Adolescents have real concerns about ment can lead to decades of disdain and cyberbullying. A the place of digital technology in their single stupid ac on can be reviewed by authori es, voters everyday lives. They are sensi ve to the or employers decades later” (Unicef, 2017, p. 94). tensions created when their desire to Misuse or disclosure of personal informa on can also in- engage online has to be weighed against crease the risk of being cyberbullied or of engaging in self- their need to protect themselves, their depreca ng behaviors, which can have a significant impact responsibili es to themselves and on an individual’s psychological and physical health, poten- others, and the responsibili es of adults ally decreasing life sa sfac on and increasing depression to help them live and grow well in the and drug and alcohol use among the vic m. Marginalized digital age” (Nyst, 2017, p. 83). groups face even greater risks of online harassment based on their race, ethnicity, or sexual orienta on. A 2013 study Digital baggage, or the digital footprint that has accumulated of 5,907 internet users in the United States age 13 to 18 since birth, can cause addi onal risks for adolescents as they found that those who self-iden fied as lesbian, gay, bisex- move into adulthood where “over a quarter (27%) admit to ual, or transgender were dispropor onately at risk of online there being ‘inappropriate’ photos of them online, while al- sexual harassment (Mitchell, Ybarra, & Korchmaros, 2014). most half (46%) say that there are photos on the Internet of them that they wish they could remove” (AVG Technologies, 2015). As kids transi on to being working adults, the privacy risks con nue to cause concern both at home and at work, where “of those that agreed social media has eroded their privacy at work, nearly a quarter (24%) now avoid pos ng
6 PRIVACY RISKS AND HARMS privacy.commonsense.org What type of personal • “Preferences” describes informa on about a person’s unique opinions, interests, or inten ons that direct their informa on is at risk? ac ons.
The tremendous amount of personal informa on collected from and about kids by applica ons and services includes a External informa on wide range of internal, external, financial, social, technologi- cal, and poli cal informa on. The more personal informa on • “Iden fica on” describes informa on that uniquely collected from, and about, a child or student, the greater the iden fies a specific person based on their characteris- content, contact, and conduct risks and magnitude of any cs such as their name, unique iden fier, photograph, resul ng harm. or biometric data.
Personal informa on is generally defined as any informa- • “Ethnicity” describes informa on about a person’s ori- on that iden fies, relates to, describes, is capable of be- gins or lineage such as their race, na onal origin, or lan- ing associated with, or may reasonably be linked, directly guages spoken. or indirectly, with a par cular child, student, or household. • “Sexual iden ty” describes informa on about a per- This means that when a kid downloads an app or logs in son’s gender iden ty, sexual orienta on, preferences, to a service and provides their informa on, that informa- proclivi es, and history with their partners. on is likely personal informa on that could be used to cause them inten onal or uninten onal harm in the future. • “Behavioral informa on” describes a person’s knowl- In addi on, the nature of collec ng personal informa on edge, traits, personality, and ac vi es that direct their from kids, through shared technology devices and applica- ac ons. ons and services used at home and in the classroom, eas- • “Demographic” describes informa on about a person’s ily allows for the associa on of a specific child to a house- characteris cs that they share with a group of other hold and their parents and a student to their school, class- persons such as age, race, gender, educa on, and geo- room, and educator (Christl, 2017). This can lead to in- graphic loca on. forma on asymmetry, where companies, organiza ons, and governments through their data-collec on, data-associa on, • “Health” describes informa on about a person’s medi- and data-recombina on processes can have be er in mate cal condi ons or health care records including informa- knowledge about a kid’s social and emo onal behaviors and on about their physical and mental health, drug tests, idiosyncrasies than the kid or parent/teacher does. disabili es, fer lity, family history, blood type, DNA, biometric data, surgeries, and drug prescrip ons. The following list of categories of personal informa on rep- resent details that may be misused, used to target or bully, • “Physical characteris cs” describes informa on about or used to manipulate a kid’s behavior, poten ally increasing a person’s unique physical characteris cs such as their the content, contact, and conduct risk exposure. This addi- height, weight, age, hair and eye color, skin tone, and onal risk exposure increases the likelihood of inten onal or gender. uninten onal harm if the respec ve personal informa on is not adequately protected. Due to these increased risks and Financial informa on poten al harms, there is an increased need to ensure that in- forma on collected is not used, or not disclosed in a context • “Accounts” describes informa on that iden fies a per- other than the context or purpose for which it was collected. son’s financial accounts such as their credit or debit card number, bank account number, or other monetary Internal informa on account number.
• “Property” describes informa on about tangible and in- • “Knowledge and beliefs” describes informa on about tangible things a person has rented, borrowed, licensed, what a person knows or believes, such as their religious or owned/owns such as a car, a house, land, digital beliefs or their private thoughts. items, or personal possessions. • “Authen ca on” describes informa on used to authen- • “Transac ons” describes informa on about a person’s cate a person with an applica on or service with financial ac vi es such as their purchases, sales, credit, something they know, such as a username and pass- income, loans, taxes, and spending habits. word or an answer to a secret ques on.
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 7 • “Credit” describes informa on about a person’s repu- tent, such as a mobile device, tablet, laptop, personal ta on with lending sources such as their earning po- computer, TV, or internet-connected smart device. ten al, spending data, and history paying back money • “So ware” describes informa on about which operat- owed to others through credit reports, credit scores, ing systems and versions of applica ons and services credit availability, and credit worthiness. a person uses, such as games and apps for communi- ca on, browsers to access websites, apps for collabo- Social informa on ra on with friends, services for content crea on, and services that deliver media entertainment or news.
• “Professional” describes informa on about a person’s • “Subscrip ons” describes informa on about the free or professional job or career such as their job tle, salary, paid plans for services a person uses to access technol- work history, subordinates, yearly evalua ons, refer- ogy, such as a broadband internet plan, mobile voice ences, interviews, cer fica ons, complaints, and disci- and data plan, television channel plan, or so ware ac- plinary ac ons. cess subscrip on plan.
• “Criminal and civil” describes informa on about a per- • “Tracking” describes informa on about the history of son’s criminal ac vity such as arrest records, trial pro- the content a person accesses and includes the device ceedings, convic ons, proba ons, and pardons as well or so ware they access the content from, what content as civil lawsuits with se lements and details of the dis- they accessed before and a er that content, at which pute. loca on they accessed it, when they accessed it, how • “Educa on” describes informa on about a person’s ed- long they accessed it, how they interacted with it, how uca onal records as a student such as their contact in- much they paid for access, who else interacted with it, forma on, family contact informa on, schools, teach- and what other content is connected to it. ers, classes, subjects, grades, assignments, and scholar- ships, as well as disciplinary ac ons taken against them. Poli cal informa on • “Public” describes informa on about a person’s public iden ty such as their character, reputa on, social sta- • “Speech” describes informa on about the type and con- tus, marital status, religion, poli cal affilia on, interac- tent of poli cal messages or opinions a person provides ons, and community affilia ons. to others and that is meant to persuade, influence, or polarize their knowledge or beliefs. • “Family” describes informa on about a person’s fam- ily rela onships such as family structure, marriages, di- • “News” describes informa on about the type and vorces, adop ons, siblings, offspring, and inheritance. source of poli cal content or opinions that a person receives and that is meant to inform their knowledge • “Social network” describes informa on about a per- or beliefs. son’s friends or connec ons that include their associa- ons, their group memberships, and the history of the • “Misinforma on” describes informa on about the ve- connec ons. racity and source of poli cal content or opinions that a person receives and that is meant to persuade or influ- • “Communica on” describes informa on about a per- ence their knowledge or beliefs. son’s messages or communica ons to others such as voice recordings, electronic mail correspondence, at- • “Contribu ons” describes informa on about monetary tachments, and instant messages sent through other gi s or other forms of value a person provides to a par- technology pla orms. Informa on about end-to-end cular candidate, poli cal party, ballot measure, or po- encrypted communica ons may s ll describe commu- li cal ac on fund. nica on metadata such as the pla orm used to send a • “Ac vism” describes informa on about the type and message, the me the message was sent and received, content of poli cal ac vi es a person par cipates in the iden fica on of the sender and receiver, and the such as rallies, marches, protests, demonstra ons, can- amount of message data sent and received. didate events, signature gathering, and civil disobedi- ence. Technological informa on
• “Devices” describes informa on about the devices with unique iden fiers a person uses to access digital con-
8 PRIVACY RISKS AND HARMS privacy.commonsense.org stolen physical possessions again, personal informa on of a PRIVACY HARMS kid lost in a data breach is o en sold or recombined with other informa on about that kid by third par es who use What are the harms? the informa on to target adver sements or steal the iden- ty of that kid, their parents, or their educators, for the rest Even though children and students might encounter online of their lives. privacy risks similar to those of parents and educators, they can experience very different outcomes in terms of harms You can turn off a non-age-appropriate (Fau & Moreau, 2018). It is important for parents and educa- movie, but you can’t turn off an app that tors to understand both the privacy risks, which include the doesn’t respect privacy. selling of data, third-party marke ng, behavioral adver sing, third-party tracking, and the crea on of adver sing profiles, Companies and schools may use products that collect ex- as well as the privacy harms, which include loss of a en- tremely detailed and sensi ve informa on about children on, poor mental health, broken rela onships, and a threat and students, including family financial informa on, health in- to our democracy (“Privacy Program,” 2019; Center for Hu- forma on, real- me loca on, biometrics such as palm prints mane Technology, 2019). Once a kid downloads an app or or fingerprints, Social Security numbers, and behavioral and logs in to a service that does not respect their privacy, it’s disciplinary records. There are immediate risks to the collec- o en too late to protect their privacy and minimize their fu- on and storage of this sensi ve personal informa on, such ture risk of exposure or the impact of any harms. If a kid as the disclosure of students’ behavioral or special educa on provides their personal informa on to a product that does records, which can lead to reputa onal and emo onal harm not respect privacy, then that kid’s informa on can be col- and poten al bullying (Nyst, 2017, p. 9). lected, used, and shared with third par es who can con nue Research indicates “whereas in previous genera ons, chil- to use that kid’s data to influence their decisions for the rest dren being bullied could escape such abuse or harassment of their lives. Even if a kid deletes their data or removes the by going home or being alone, no such safe haven exists for app, the harm does not go away; it only stops the spread children in a digital world. Carrying a mobile phone, laptop or of addi onal harm with that applica on or service. The po- other connected device means that texts, emails, chats and ten al harm keeps spreading. The impact of dele ng a kid’s social media posts can arrive any me, day or night. And on- data is only rela ve to the applica on or service to which line bullying carries on, spreading widely among peers and they provided their data in the first place, because dele on inflic ng reputa onal harm whether the child is online or does not stop the informa on from spreading to third par- off” (Unicef, 2017, p. 74). If a child’s real- me loca on is ob- es with access to the data. Children and students typically tained by a “bad actor,” such as decep ve adver sers, abu- have no ability to protect their privacy or exert control over sive (ex)partners of a parent, or child groomers, they could how third par es use their data once it has been collected pose direct physical and safety concerns. Even if the number and shared (Carpenter v. United States, 2018). of children suffering severe harm is rela vely low, when harm does occur to a child, according to review of evidence in this A er a kid’s data has been shared with area, its impact on the child can be significant and jus fies the third-party companies, those same alloca on of substan al resources and a en on to prevent companies can s ll use their own copies the harm (Slavtcheva-Petkova, Nash, & Bulger, 2015). First of of the data for their own purposes. all, simply coun ng the number of children who report that they personally have been bullied does not capture all of the This harm is further exacerbated in the event of data breach. harm that occurs as the result of online bullying. While the For example, if a kid’s personal informa on is publicly dis- number of children who are bullied, especially the percent- closed to third par es in a data breach, then it is difficult or age of those affected who report being bullied, may be small impossible to stop it from being shared indefinitely, and there as a percentage of overall users, the culture of online bullying is no simple way to stop the poten al harm from spreading, hurts everyone. Bullying and other nega ve communica on even if the kid were to delete all of their data from the com- online affect children who see it, hear about it, read about it, pany that had the data breach a er the fact (“K-12 Cyber- and live it. It is a culture of fear. Parents and educators need security Resource Center,” 2019). In addi on, the poten al to be er understand the privacy risks and poten al harms harm to a kid is propor onal to the type and amount of sen- of applica ons and services so they can make informed de- si ve personal informa on disclosed. More sharing of per- cisions about which apps their children and students should sonal and in mate details about a kid amplifies both the po- use before they download or log in to them, not a erward, ten al and magnitude of inappropriate use and harm. Unlike because the harm will have already occurred. in a home burglary where you will not likely ever see your
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 9 Social and emo onal harms commercializa on of kids’ private data can result in harm to children. In a now famous example of embarrassing disclo- A kid’s informa on can also be misused to create lifelong sure, Target analyzed a teenage girl’s data and decided she social and emo onal harm. Sensi ve personal informa on was pregnant and disclosed this informa on to her father about a child’s anxie es, fears, secrets, knowledge, and be- (Hill, 2012). liefs were tradi onally kept private in a diary or shared only Research shows that behavioral adver sing that uses per- in confidence with very close family, friends, or adults they sonal informa on to target online ads to specific behaviors, explicitly trusted such as their parents or educators. This sen- as well as other adver sing techniques, can contribute to the si ve personal informa on o en defines an individual’s in- growing commercializa on of childhood (Palfrey, Gasser, & nermost secrets and sense of self-worth and security (Nyst, boyd, 2010). For example, disclosure of sensi ve personal in- 2017, p. 91). forma on could cause social harms such as loss of a parent Sharing this informa on in confidence can help build that or educator’s money or employment status if the informa on kid’s founda on of trust and safety. Sharing this informa on contained nega ve remarks about that parent or educator’s online, however, turns social media, which by its name sug- company, supervisors, or colleagues. For example, the for- gests a public func on, into the publishing of a private diary. mer head of a private preparatory school in Miami, Florida, There are two basic methods that place the private diary in lost an $80,000 confiden al discrimina on se lement a er public hands: One, children with li le understanding of how his daughter boasted about it on Facebook (Stucker, 2014). privacy or technology work “voluntarily” share this informa- In addi on, a kid may experience reputa onal harm or cy- on online, expec ng just a friend or small circle of close berbullying if their sensi ve personal informa on is inten- acquaintances to see the material. Two, sensi ve personal onally or uninten onally disclosed to others at their school; informa on can be collected knowingly or unknowingly by a that informa on can include their private messages, educa- company through that kid’s use of an applica on or service on records, in mate details, or explicit photos. A research at home or in the classroom and used at any me to exploit survey of parents in the U.S. found that 26% reported that their vulnerabili es. their child was a vic m of cyberbullying in 2018 (Cook, We cannot, nor should we, eliminate all risks, but the most 2018). Inadvertent disclosure of sensi ve informa on to a vulnerable are those most exposed to risk and therefore the kid’s peer groups could lead to cyberbullying and extreme so- most likely to suffer the harms. It is cri cally important to cial changes such as social or emo onal learning problems in understand why risk can translate into actual harm for certain the classroom, poor learning outcomes with students repeat- children and not for others. Some children may be unaffected ing grades, disciplinary ac ons, or even expulsion (Tokunaga, by luck of the draw, in that their data isn’t chosen or used 2010). inappropriately. Other children may be vic mized, especially For example, for months, 12-year-old Mallory Grossman if they fall into a par cularly vulnerable category as discussed received taunts in text messages, Instagram posts, and in this report’s sec on “Who is vulnerable to the risks?” (pg. Snapchats from classmates that said “why don’t you kill 2). yourself?” The taunts, her parents say, took a toll on the lively young cheerleader and gymnast. At school, Mallory’s “It opens our eyes to the underlying grades deteriorated, and at home, she complained of con- vulnerabili es in the child’s life that can stant headaches and stomach pain. Mallory begged to stay place him or her at greater risk in the home from school, and then on June 14, 2017, she took digital age. By be er understanding and her own life (Schmidt, 2017). There are also real-word con- addressing these vulnerabili es, we can sequences for the perpetrators of cyberbullying. In another be er protect children both online and example, more than a dozen students were expelled from offline, and be er enable them to enjoy Boulder Preparatory High School in Boulder, Colorado, af- the opportuni es that come from being ter being involved in pos ng rape memes, messages cham- connected in the digital age” (Unicef, pioning “white power,” and comments to their classmates 2017, p. 71). about wan ng to kill black and Jewish people in a group called the 4th Reich Official Group Chat on Facebook (Levin, 2016). If sensi ve personal informa on is used to target children, students, or their parents or educators with behavioral adver- In the American Sex and Tech survey from 2008, 39% of 13- sements, those targeted adver sements could cause harm to 19-year-olds reported that they had sent or posted sexu- to a kid if the informa on in the message was also disclosed ally sugges ve messages, 20% reported that they had sent or to that kid’s family, friends, or peer groups. We don’t need posted nude or seminude photos of themselves, and 38% re- a data breach or a bad actor for things to go wrong. Simple ported that the act of sex ng someone made da ng or hook-
10 PRIVACY RISKS AND HARMS privacy.commonsense.org ing up with that person more likely (Kosenko, Luurs, & Binder, which can produce chilling effects on forms of ac on or ex- 2017). Disclosure of this sensi ve informa on either inten- pression (Shaw, 2017; Christl, 2017). onally or uninten onally could cause irreparable behavioral, Children and students who change their minds and wish to emo onal, physical, or social harm to a kid that could result in keep their sensi ve personal informa on private and who sudden changes to their mood or personality and in extreme choose to opt out of providing their personal informa on cases lead to self-harm, such as cu ng, or increased risk of to apps that do not respect their privacy, or who refuse to suicide (Nixon, 2014; Kowalski & Limber, 2013; Centers for opt in and subject themselves to surveillance and monitor- Disease Control and Preven on, 2018). ing technologies, are o en seen as rejec ng social norms and are excluded from school and peer ac vi es. However, “Strong evidence shows that girls face these choices mean that students may not be able to take much greater pressure to send sexually advantage of educa onal technology used in the classroom explicit images and suffer much harsher and at home to complete assignments, which may put them judgements when those images are at risk for lower test scores and at a learning disadvantage shared beyond the intended recipient” compared to their peers. At the same me, children and stu- (Livingstone & Mason, 2015, p. 10; boyd, dents have ever fewer op ons to resist the power of this data Ryan, & Leavi , 2011). ecosystem; op ng out of pervasive tracking and profiling has essen ally become synonymous with op ng out of much of Sexual blackmail has affected very young teens. For exam- modern life (Christl, 2017). This social harm can result in in- ple, “when Amanda Todd, a Canadian adolescent, was about creased social isola on and a cycle of exclusion from peer 13, a man she met in a video chat room convinced her to groups that can lead to anxiety and depression. Ul mately, expose her breasts on camera. He captured the image and peer pressure to use normalized surveillance technologies in used it to blackmail her, threatening to send the image to her order to par cipate in social ac vi es results in children and friends and family. She ignored the threat and over the next students reluctantly using products that do not respect their two years was subject to bullying (both online and offline), privacy to try to conform to societal norms. harassment and physical assault. Despite her efforts to es- For example, kids have reported that when they know all cape the torment—she moved both schools and ci es—the their online ac vi es are being monitored at home or at a acks con nued, both by the online offender and by her school by surveillance technologies, they change their be- classmates. During this me, she struggled with depression, havior in response and are less likely to engage in conver- drug and alcohol abuse, isola on, loneliness and self-harm. sa ons or take risks to learn something new or explore con- Two years later, in October 2012, at 15, Amanda commi ed troversial ideas that may get them into trouble. As a result, suicide” (Nyst, 2017, p. 74). While for boys, possessing and children and students have been found to be less likely to exchanging explicit images of girls adds to their reputa on engage in cri cal thinking, par cipate in poli cal ac vism, and status among peers. However, for girls, par cipa ng in vote, or ques on news or authority figures (Brown & Pecora, sending explicit images or “sex ng” raises concerns about the 2014). Cambridge Analy ca, a data-analysis firm, exfiltrated poten al risks to their sexual reputa on among peers (e.g., data on 50 million Facebook users in early 2014 to build being called a “slut”) (Ringrose, Harvey, Gill, & Livingstone, a system that could profile individual U.S. voters to target 2013). them with personalized poli cal adver sements and influ- ence their cri cal thinking and vo ng behavior (Cadwalladr Surveillance harms & Graham-Harrison, 2018). Research indicates “Cambridge Analy ca sought to iden fy mental and emo onal vulner- Parents and schools that use technology to monitor a kid’s abili es in certain subsets of the American popula on and daily ac vi es can automa cally collect all the informa on a worked to exploit those vulnerabili es by targe ng informa- kid generates on applica ons or services, which can include on designed to ac vate some of the worst characteris cs sensi ve personal informa on. This collec on prac ce can in people, such as neuro cism, paranoia and racial biases” serve to normalize surveillance technologies for children and (Wri en statement to the United States Senate Commi ee students at a young age and can change the way they regard on the Judiciary, 2018). Surveillance technologies with mar- their private informa on. Children with resul ng lower ex- ke ng interests have also steered the development of digi- pecta ons of privacy may then assume that all technologies tal networks toward maximizing their consumer surveillance are engaged in the same type of surveillance ac vi es, which capaci es, which fer lized the soil for poli cal manipula on can lead them to download apps that also do not respect (Crain & Nadler, 2017). their privacy (Unicef, 2017, p. 9). In addi on, research shows that when people know they are being constantly monitored, they change their behavior in response to the surveillance,
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 11
Ideological harms informa on may not even be able to iden fy it as influen al in their choices (Menn, 2018). Rather than simply collect personal informa on to match consumers with products that fit their exis ng interests, mass consumer surveillance technologies have led to so- Who poses a risk of harm to phis cated efforts to modify behavior, engineer consumer kids? habits, and intervene upon in mate decision-making pro- cesses (Crain & Nadler, 2017). Personal informa on col- Research shows that kids value their privacy online and they lected from children and students and their parents over me see risks to their privacy coming from both outside their per- can be used to influence their decision-making processes sonal circle, such as from companies and governments, and now or in the future, which ul mately causes harm to our from within their own circle, such as from overprotec ve par- society and democracy when it results in behavior modifica- ents, nosy parents, and parents, friends or siblings who spy ons that change, for example, vo ng preferences and elec- (Third, Bellerose, Dawkins, Kel e, & Pihl, 2014). on outcomes (Epstein & Robertson, 2015). Poli cal par c- ipa on can be broken down into units of specific behaviors. It’s at this granular level where nudges, emo onal triggers, First- and third-party companies and carefully designed choice architecture can exert signifi- cant influence at cri cal steps (Crain & Nadler, 2017). A kid’s personal informa on is inherently at risk when it’s col- lected by a company that they have a first-party rela onship In a 2010 study published in Nature, Facebook reported an with, such as an app the kid downloads or a website they experiment involving over 61 million Facebook users who log in to. These companies may use that same personal in- were randomly selected to see different types of messages forma on to indirectly influence or modify the behavior of or no message about vo ng in their Facebook News Feed the kid or directly influence the behavior of the kid’s par- on Elec on Day. The authors of the study found that the ent or educator through first- and third-party marke ng and Facebook messages likely brought an extra 340,000 people adver sing. This influence can be posi ve in nature, in that to the poll that day (Talbot, 2012). However, Facebook also useful informa on about products is transmi ed. S ll, in the used “distributed targeted poli cal ads apparently sponsored case of marke ng to children, we need to be careful in what by Russian opera ves. Most of the ads made no reference to we allow children to see and to buy. In-app purchases are specific candidates. Rather they appeared to focus on ampli- especially problema c as kids may not realize they are buy- fying divisive social and poli cal messages across the ideolog- ing with real money rather than playing a game. Moreover, ical spectrum—touching on topics from LGBT ma ers to race many of the privacy protec ons afforded to children under issues to immigra on to gun rights” (Crain & Nadler, 2017). 13 years of age with the Children’s Online Privacy Protec on Poli cal polariza on of social issues through ads can be used Act (COPPA) do not apply to children over 13 (Children’s On- to make individuals ques on their strongly held knowledge line Privacy Protec on Act, 2012). Teens’ personal informa- or belief systems, which can lead to voter suppression if in- on is at an even greater risk of misuse, because it can be dividuals feel they no longer understand the issues or that collected and disclosed to third par es without their knowl- their vote doesn’t ma er. edge or their parent’s consent (Felt & Robb, 2016). “If data-driven targe ng and behavioral First-party companies then may use the informa on gath- science can be used to increase voter ered by the direct rela onship and share that informa on turnout, it can also be used to suppress with third-party companies that have no rela onship with it. Here, the task is to iden fy marginal the original underage user. Children or trusted adults can be influenced to purchase other products from that first- voters leaning toward an opponent and party company or, for the children and students, products figure out what behavioral interven on from other third-party companies. With good inten ons of might nudge them not to vote” (Crain & keeping individual privacy intact, many companies mone- Nadler, 2017). ze data collected from children and students by so-called de-iden fying or anonymizing the personal informa on and Informa on that affects voter turnout could be really simple, building personality or user profiles of children and students like real or fake informa on about the weather, the lines at to sell to other third-party companies or to license the data the polls, and/or the possible outcome of the elec on. None for targeted adver sing (General Data Protec on Regula on, of this has anything on its face that has to deal with choosing 2018). A kid’s de-iden fied or anonymized data may also be a candidate but might be even more effec ve because it is reiden fied or aggregated later by any of the third par es subtle and less likely to raise alerts. The recipients of such along the line once the data has le the care of the first-party
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 13 company with a direct legal or contractual obliga on to main- Organiza ons tain the privacy of kids’ data. O en, data provided to one ap- plica on or service that is not considered personal informa- A kid’s personal informa on may also be used by organiza- on may s ll be shared and combined later with other data ons such as educa onal ins tu ons or employers for dis- from a par cular kid from another third-party applica on or crimina on. However, the Civil Rights Act of 1964 prevents service in order to create valuable personal informa on that discrimina on in educa onal facili es and public workplaces can be used to re-iden fy that kid or link them to separate for specific protected classes of personal informa on. Under data sets (Nyst, 2017). This reiden fica on of a kid’s informa- that act and other federal and state an discrimina on laws, a on is o en performed without the child, student, parent, or person may not be discriminated against based on any of the educator’s knowledge or consent. following: age, pregnancy status, na onal origin, race, eth- nic background, religious beliefs, or sexual orienta on (Civil A kid’s personal informa on may also be used by companies Rights Act of 1964). that provide banking, housing, or insurance to discriminate against that kid or their parents (Dreyfuss, 2019). Histori- Despite these laws, a kid’s personal informa on that is not cally, companies used a now-illegal prac ce called “redlining,” protected could s ll be used as a proxy for discrimina on if which used the personal informa on and ZIP codes of com- it provides informa on about themselves, their parents, or muni es of color to deny them access to housing or banking their educators. Organiza ons can easily purchase personal services. However, personal informa on provided from a kid informa on about children, students, and their parents or about themselves or their parents such as criminal history, educators from third-party companies and data brokers. Al- credit history, educa on level, economic status, or even ZIP though discrimina on can exist within all types of classes, code could s ll be used by companies to deny them bank there are certain classes of personal informa on that are not accounts, car loans, access to credit, or rental applica ons or protected under an discrimina on laws: result in the charging of higher insurance or interest rates. • ci zenship status Companies today use as much personal informa on as they can collect about a child or their parent in their proprietary • credit history algorithm decision-making process to determine whether to • criminal history accept or deny services, or even to determine whether adver- sements will be displayed for be er opportuni es. These • economic class decisions are o en based on thousands of data points and • educa on level weighted factors, which makes determining whether illegal discrimina on occurred nearly impossible. • membership in organiza ons
The U.S. Department of Housing and Urban Development • physical characteris cs determined that the social media giant Facebook violated the • ZIP code Fair Housing Act because it “allows landlords and home sell- ers access to adver sing tools that limit which prospec ve Regardless, organiza ons subject to the Civil Rights Act can buyers or tenants can view certain online ads based on race, s ll serve or not serve adver sements based on protected religion, sex, disability and other characteris cs” (Booker, classes of personal informa on if the adver sements are not 2018). For example, Linda Bradley of Franklin County, Ohio, discriminatory. Organiza ons can s ll use personal informa- lost her job at a call center and had been searching for work on not protected under an discrimina on laws to deny chil- through Facebook, but Bradley and other female members dren or their parents acceptance into higher educa onal pro- of the union discovered they were rou nely denied the op- grams, deny children internship opportuni es, deny their par- portunity to receive job ads and recruitment opportuni es ents or educators employment opportuni es, or offer be er on Facebook that “similarly situated male Facebook users ... pricing to some groups or individuals. received” (Tiku, 2018). Companies have access to more personal informa on about Governments children and their parents than ever before. Technological advancements have enabled companies to use that data A kid’s personal informa on is also at risk when it’s collected to discriminate against vulnerable popula ons by engaging by or shared with the U.S. government. Any personal infor- adver sing-targe ng technology to exclude them from see- ma on a kid shares with a company or organiza on also can ing be er opportuni es in or informa on about housing, be shared with the government at any point they request it jobs, educa on, or credit (Booker, 2018). for the rest of that kid’s life. It is common for a company’s poli- cies to allow them to easily share a kid’s informa on with the government for any purpose that they believe is necessary
14 PRIVACY RISKS AND HARMS privacy.commonsense.org to protect the safety of their company or their users. How- Since the data is collected by a private company rather than ever, some companies’ policies state they will only share a the government directly, important legal safeguards estab- kid’s personal informa on with the government if a govern- lished to protect ci zens from government overreach may be ment agency presents a subpoena for the informa on based completely ignored. It is possible that the only thing protect- on probable cause that a crime has been commi ed and the ing our privacy at this point in me is that the sheer quan ty kid’s informa on must be disclosed as part of the inves ga- of video images gathered by these cameras and other data- on (Cardozo et al., 2018). collec on methods is too difficult to use, but that is just for now. “If governments are able to link Personal informa on from a kid could even be used against individual profiles with data intercepted them or their parents by a foreign government to persuade by mass surveillance, as many believe them to act as an undercover agent or informant against oth- feasible, this would allow authori es to ers to avoid prosecu on. Other foreign governments that build and maintain records of children’s obtain sensi ve personal informa on from children or stu- en re digital existence” (Nyst, 2017, dents could even use that informa on to blackmail or per- p. 15). suade them or their parents to engage in espionage or trea- son against their own government to avoid prosecu on or However, governments do not always need to request that even worse consequences by their own government if the companies or organiza ons provide them with personal in- sensi ve personal informa on were publicly disclosed. In ex- forma on from children or students that they have collected, treme cases in some countries around the world, a kid’s per- because many governments already implement their own so- sonal informa on that has been collected or shared with a phis cated data-collec on surveillance technologies that au- repressive foreign government about their poli cal ac vism, toma cally intercept and collect personal informa on from homosexual iden ty, or atheist religious beliefs could lead its ci zens and ci zens of other countries for na onal secu- to the most extreme type of harms against them or their rity purposes without their knowledge. Research indicates parents such as imprisonment, public stoning, or the death “governments can collect vast amounts of online personal penalty (Ohlheiser, 2013; Lamb, 2019; Bearak & Cameron, data on children, a type of surveillance largely unimaginable 2016). in the pre-internet era. O en neither lawful nor publicly ac- knowledged, mass surveillance now forms a key part of na- Bad actors onal security efforts in many countries. Not only does it un- dermine basic no ons of privacy, it also threatens other basic Children and student’s personal informa on can also be used human rights, such as freedom of expression, and opens the by other individuals with bad inten ons to cause them inten- door to poten al abuses of state power” (Brown & Pecora, onal physical or emo onal harm. Family members, rela ves, 2014, p. 202). family friends, acquaintances, neighbors, or even strangers can collect and use personal informa on from children and Facial-recogni on and other students to gain specific and in mate knowledge about them image-recogni on tools will soon be and use that knowledge for their own personal gain and ob- able to make this data much, much more jec ves. Research indicates that children and teens consider useful to law enforcement and other mee ng a stranger online an opportunity to meet new peo- government en es, but it is s ll prone ple or even develop roman c rela onships, while parents to errors (Singer, N., 2018; Sengupta, see mee ng a stranger online as one of the most danger- 2013). ous things that could happen to their child on the internet. (Phyfer, Burton, & Leoschut, 2016; Mascheroni & Cuman, Depending on the type of personal informa on disclosed to 2014). Research examining studies between 1990 and 2016 a government and that government’s laws, there is a wide found that approximately 20% of youth had been exposed to range of risks and possible harms that could result. Many gov- unwanted sexual content online and 11% experienced un- ernments when inves ga ng a crime may collect and analyze wanted online sexual solicita on, with 25% of youth report- huge volumes of personal informa on about its ci zens, such ing being “extremely” bothered by these experiences (Madi- as mobile app and device phone calls, messages, photos, ge- gan et al., 2018). In mate knowledge learned about a child oloca on histories, emails, and even health informa on such can give bad actors a power differen al that allows them to as familial DNA that could be used to prosecute related fam- influence a kid’s ac ons as well as influence their family or ily members. Governments may also rou nely collect data friends. through a system of contracts with private companies, such as asking a private company to install cameras on city streets.
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 15 “Advances in technology allow offenders to remain anony- Children’s and students’ informa on can also be abused by mous, cover their digital tracks, create false iden es, pur- fraudsters peddling fake scholarship opportuni es. Social Se- sue many vic ms at once, and monitor their whereabouts. curity numbers and other background informa on can be The increased use of mobile devices and greater access to stolen and used by iden ty thieves to open up new lines broadband internet have made children more accessible than of credit, crea ng serious problems for children and parents ever through unprotected social media profiles and online who o en don’t discover they have been vic mized un l a game forums. Offenders o en begin grooming their vic ms number of years have passed; the ease of obtaining a child’s on these pla orms, where they gain a child’s a en on or informa on plus delayed discovery, combined with clean trust, before moving the communica on to video- and photo- credit reports, are reasons children are par cularly suscep - sharing pla orms, which can lead to content-driven or finan- ble to iden ty the . Research has shown that children are 35 cially driven extor on or mee ng offline” (Nyst, 2017, p. 76). mes more likely to be vic ms of iden ty the because they don’t have a credit history and their Social Security number Extremist groups can also use digital and communica on isn’t ac ve, according to the Division of Consumer Protec- technologies to collect personal informa on about kids in or- on of the New York Department of State (Singer, P., 2018). der to make contact, radicalize them, and persuade them to take part in unhealthy or dangerous ac vi es, such as striking targets with whatever weapons are available, such as knives How parents and educators or crude bombs (Faiola & Mekhennet, 2017). For example, personal informa on about a kid’s first and last names, their can make a difference parent or educator’s names, their birthday, and their home address could be used by non-parents to impersonate that Privacy is complicated, and protec ng kids online, regardless kid’s parent, family member, or educator to persuade oth- of the par cular privacy risks and harms they face, requires a ers to contact or give over custody of the kid in order to holis c and coordinated response that takes into account all kidnap them or cause them harm. The more personal infor- the circumstances of a child’s life. Risk is specific to each kid’s ma on a bad actor can collect about a kid the more repu- unique situa on, and safeguards and protec ons will need to ta onal influence and persuasion they have over others en- be adjusted as kids come of age. Understanding the risks will trusted with the care of those children and students (Bilich, also help minimize the poten al harms and maximize the pos- n.d.). For example, the recent documentary Abducted in Plain i ve outcomes. Parents, teachers, companies, organiza ons, Sight showed how a sexual predator collected and used per- governments, and kids themselves all have a part to play in sonal informa on over me to become an acquaintance and keeping kids safe. These recommenda ons can make a dif- torment an Idaho girl, and her family, for years (“Abducted in ference in outcomes for kids: Plain Sight,” 2017). 1. Read high-quality evalua ons about which online tech- Bad actors can also engage in physical or emo onal abuse nologies protect kids’ privacy. As parents, educators, and such as physiological abuse, blackmail, kidnapping, child mo- consumers, we can help encourage companies to make lesta on, exploita on, or sex trafficking. Research indicates changes in how they collect and use personal informa on that “even as informa on and communica on technology from kids by only purchasing products for kids that protect has made it easier to share knowledge and collaborate, so, their privacy and avoiding products that do not. Parents and too, has it made it easier to produce, distribute and share sex- educators can also use the easy-to-understand privacy eval- ually explicit material and other illegal content that exploits ua ons from Common Sense. Privacy evalua ons include an and abuses children. Such technology has opened new chan- overall score, display er risks, and summarize privacy con- nels for the trafficking of children and new means of con- cerns to help parents and educators make informed choices cealing those transac ons from law enforcement. It has also about the products they use at home and in the classroom made it far easier for children to access inappropriate and po- (“Privacy Program,” 2019). With the Common Sense Privacy ten ally harmful content—and, more shockingly, to produce Evalua ons, anyone can confront privacy concerns before such content themselves” (Unicef, 2017, p. 8). Applica ons they start. That’s why the Common Sense Privacy Program and services that collect and share specific personal informa- was created: to champion child and student privacy and to on about children and students such as their geoloca on support parents, educators, schools, and communi es on a or status update informa on over me present a unique risk. path toward a more secure digital future for all kids. If this specific personal informa on is made publicly avail- 2. Require commitments to safeguard kids’ privacy. Parents able or accessed directly or indirectly by bad actors, the data and educators should be able to harness the true power of could be used to stalk a kid over me and learn in mate de- informa on and communica on technologies to benefit all tails about their home and school addresses, rou nes, and children, especially the most disadvantaged children, while friends or even the best me of day to confront them alone. at the same me limi ng the harms to protect those chil- dren who are most vulnerable (Nyst, 2017, pp. 11, 84). How-
16 PRIVACY RISKS AND HARMS privacy.commonsense.org ever, digital technologies pose significant risks to children’s of verbal and facial clues to give meaning and its poten al safety, privacy, and well-being. The harms that many chil- for anonymity (Nyst, 2017, p. 129). Social-emo onal learn- dren already face offline can make already vulnerable chil- ing and the teaching of empathy helps develop kids’ online dren even more vulnerable (Nyst, 2017, p. 8). Today’s surveil- resilience and helps to diminish online abuse and hateful lance technology and data footprints allow for children’s data language. Check out the Common Sense Digital Ci zenship to be combined and reused in both amazing and alarming Curriculum for more resources to take on mely topics for ways. Kids deserve a much greater commitment by private school communi es, support teachers with improved class- companies, organiza ons, and governments to protect their room tools, and prepare kids to take ownership of their digital data and a shared agreement not to misuse or exploit kids’ lives (“Digital Ci zenship,” 2019). data for their own purposes. Prac ces that take advantage of kids’ suscep bility, developing mental capacity, and inability to discern whether adver sing messages are truthful, accu- rate, and unbiased should be off-limits to kids. Parents and CONCLUSION educators also need to teach children by example how to protect themselves from threats to their own privacy and With this report, we collect the best available informa on iden ty (Unicef, 2017, p. 11). about ways consumers can arm themselves with informa- on when choosing which technology tools to use. There Parents and educators can begin by choosing privacy- is no one-size-fits-all solu on for privacy, and so parents protec ve products and then con nue role-modeling by us- and teachers need to educate themselves with resources like ing privacy protec ons for themselves and refusing to offer ours and those offered by other trusted sources in order to their children’s informa on to social media and other prod- best understand how to minimize the risk of harms to our ucts that may use their informa on for purposes outside of youngest consumers based on the personal informa on col- the context in which the informa on was originally offered. lected from them, who has access to it, and how it is used. Ul mately, with support from government and industry, par- Our work at the Common Sense Privacy Program steps into ents and educators can make a real difference in safeguard- the space between what we need to do to protect our kids ing their kids’ privacy by ensuring that digital technologies and the technology that has permeated every aspect of their use kids’ personal informa on in ways that support the rights lives. Now is a pivotal moment in our ability to stop and ex- of the child to: amine what we have created, evaluate how the technology • keep them safe from exploita on risks, including the does and does not protect kids’ privacy, and demand be er risks of commercial or sexual exploita on and sexual products and services for our kids. abuse;
• protect and support their health and well-being;
• protect and support their physical, psychological, and emo onal development;
• protect and support their need to develop their own views and iden ty;
• protect and support their right to freedom of associa- on and play;
• recognize the role of parents in protec ng and promot- ing the best interests of the child; and
• recognize the evolving capacity of the child to form their own view, and give due weight to that view (In- forma on Commissioner’s Office, 2019).
3. Teach digital literacy and digital well-being. Kids need to understand the risks of content crea on and sharing infor- ma on online, including learning how to protect their pri- vacy and personal data with privacy se ngs (Informa on Commissioner’s Office, 2019). In addi on, kids need to un- derstand the many ways in which communica ng online is different from tradi onal communica on because of its lack
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 17 REFERENCES
Abducted in Plain Sight. (2017). Retrieved from https://www.imdb.com/title/tt3444312/
American Psychological Associa on. (2004, February 23). Television adver sing leads to unhealthy habits in children; says APA task force. [Press release]. Retrieved from https://www.apa.org/news/press/releases/2004/02/children-ads
AVG Technologies. (2015). The AVG 2015 digital diaries. Retrieved from https://www.scribd.com/document/123136168/AVG- Digital-Diaries
Bearak, M., & Cameron, D. (2016, June 13). Here are the 10 countries where homosexuality may be punished by death. The Washington Post. Retrieved from https://www.washingtonpost.com
Bilich, K. (n.d.) Child abduc on sta s cs for parents. Parents. Retrieved from https://www.parents.com
Booker, B. (2018, August 19). HUD hits Facebook for allowing housing discrimina on. Retrieved from https://www.npr.org boyd, d., Ryan, J., & Leavi , A. (2011). Pro-self-harm and the visibility of youth-generated problema c content. A Journal of Law and Policy for the Informa on Society, 7(1), 40.
Brown, D., & Pecora, N. (2014). Online data privacy as a children’s media right: Toward global policy principles. Journal of Children and Media, 8(2), 201–207. https://dx.doi.org/10.1080/17482798.2014.893756
Burke, M., & Kraut, R. (2016). The rela onship between Facebook use and well-being depends on communica on type and e strength. Journal of Computer-Mediated Communica on, 21(4), 265–281. https://dx.doi.org/10.1111/jcc4.12162
Cadwalladr, C., & Graham-Harrison, E. (2018, March 17). Revealed: 50 million Facebook profiles harvested for Cambridge Analy ca in major data breach. The Guardian. Retrieved from https://www.theguardian.com
Cardozo, N., Crocker, A., Gebhart, G., Lynch, J., Opsahl, K., & York, J. C. (2018, May 31). Who has your back? Censorship edi on 2018. Retrieved from https://www.eff.org
Carpenter v. United States, 138 S. Ct. 2206 (2018).
Casey, B. J., Jones, R. M., & Hare, T. A. (2008). The adolescent brain. Annals of the New York Academy of Sciences, 1124, 111–126. https://dx.doi.org/10.1196/annals.1440.010
Centers for Disease Control and Preven on. (2018). Youth risk behavior surveillance — United States, 2017 (Morbidity and Mortality Weekly Report). Retrieved from https://www.cdc.gov/healthyyouth/data/yrbs/pdf/2017/ss6708.pdf
Center for Humane Technology. (2019). Ledger of harms. Retrieved from https://ledger.humanetech.com
Children’s Online Privacy Protec on Act, 16 C.F.R. Part 312 (2012).
Christl, W. (2017). Corporate surveillance in everyday life: How companies collect, combine, analyze, trade, and use personal data on billions. Retrieved from Cracked Labs website: http://crackedlabs.org/dl/CrackedLabs_Christl_CorporateSurveillance.pdf
Civil Rights Act of 1964, 42 U.S.C. § 2000d et seq. (1964).
Cook, S. (2018, November 12). Cyberbullying facts and sta s cs for 2016-2018. Retrieved from https://www.comparitech.com
Crain, M., & Nadler, A. (2017, September 27). Commercial surveillance state: Blame the marketers. Retrieved from https:// nplusonemag.com
Digital Ci zenship. (2019). Retrieved from https://www.commonsense.org/education/digital-citizenship
Dreyfuss, E. (2019, March 9). Facebook changes its ad tech to stop discrimina on. Wired. Retrieved from https://www.wired.com
Epstein, R., & Robertson, R. E. (2015). The search engine manipula on effect (SEME) and its possible impact on the outcomes of elec ons. Proceedings of the Na onal Academy of Sciences of the United States of America, 112(33), E4512–E4521. https: //dx.doi.org/10.1073/pnas.1419828112
Faiola, A., & Mekhennet, S. (2017, February 11). They’re young and lonely. The Islamic State thinks they’ll make perfect terrorists. The Washington Post. Retrieved from https://www.washingtonpost.com
18 PRIVACY RISKS AND HARMS privacy.commonsense.org Fau, S., & Moreau, Y. (2018). Managing tomorrow’s digital skills: what conclusions can we draw from interna onal compara ve indicators? Retrieved from UNESCO Digital Library website: http://unesdoc.unesco.org/images/0026/002618/261853e.pdf
Felt, L. J., & Robb, M. B. (2016). Technology addic on: Concern, controversy, and finding balance. Retrieved from Com- mon Sense website: https://www.commonsensemedia.org/sites/default/files/uploads/research/csm_2016_technology_ addiction_research_brief_0.pdf
General Data Protec on Regula on (GDPR), Recital 38, Regula on (EU) 2016/679.
Graafland, J. H. (2018). New technologies and 21st century children: Recent trends and outcomes. Retrieved from Organisa on for Economic Coopera on and Development website: https://www.oecd-ilibrary.org/education/new-technologies-and-21st- century-children_e071a505-en
Grabe, S., Ward, L. M., & Hyde, J. S. (2008). The role of the media in body image concerns among women: A meta-analysis of experimental and correla onal studies. Psychological Bulle n, 134(3), 460–476. https://dx.doi.org/10.1037/0033-2909.134.3. 460
Hill, K. (2012, February 16). How Target figured out a teen was pregnant before her father did. Forbes. Retrieved from https: //www.forbes.com
Huckvale, K., Torous, J., & Larsen, M. E. (2019). Assessment of the data sharing and privacy prac ces of smartphone apps for depression and smoking cessa on. The Journal of the American Medical Associa on Network Open, 2(4). https://dx.doi.org/10. 1001/jamanetworkopen.2019.2542
Informa on Commissioner’s Office. (2019). Age appropriate design: A code of prac ce for online services. Retrieved from https://ico. org.uk/media/about-the-ico/consultations/2614762/age-appropriate-design-code-for-public-consultation.pdf
The K-12 Cyber Incident Map. (2019). Retrieved from https://k12cybersecure.com/map
Kelly, G., Graham, J., & Fitzgerald, B. (2018). 2018 state of edtech privacy report. Retrieved from Common Sense website: https: //www.commonsense.org/education/sites/default/files/tlr-blog/cs-state-of-edtech-privacy-report.pdf
Kosenko, K., Luurs, G., & Binder, A. (2017). Sex ng and sexual behavior, 2011-2015: A cri cal review and meta-analysis of a growing literature. Journal of Computer-Mediated Communica on, 22(3). https://dx.doi.org/10.1111/jcc4.12187
Kowalski, R. M., & Limber, S. P. (2013). Psychological, physical, and academic correlates of cyberbullying and tradi onal bullying. Journal of Adolescent Health, 53(1), S13–S20.
Kross, E., Verduyn, P., Demiralp, E., Park, J., Seungjae, D. L, Lin, N., ... Ybarra, O. (2013). Facebook use predicts declines in subjec ve well-being in young adults. PLoS ONE, 8(8). https://dx.doi.org/10.1371/journal.pone.0069841
Lamb, K. (2019, April 3). Brunei brings in stoning to death for gay sex, despite outcry. The Guardian. Retrieved from https://www. theguardian.com
Lenhart, A. (2015). Teens, technology and friendships. Retrieved from Pew Research Center website: http://www.pewinternet.org/ 2015/08/06/teens-technology-and-friendships
Levin, S. (2016, October 14). Students expelled a er Facebook group calls for ‘execu on’ of Jews, black people. The Guardian. Retrieved from https://www.theguardian.com
Livingstone, S., Haddon, L., Görzig, A., & Ólafsson, K. (2011). EU kids online: final report. Retrieved from London School of Economics and Poli cal Science website: http://eprints.lse.ac.uk/39351/1/EU_kids_online_final_report_%5BLSERO%5D.pdf
Livingstone, S., Haddon, L., Görzig, A., & Ólafsson, K. (2011). Risks and safety on the internet: the perspec ve of European children: full findings and policy implica ons from the EU Kids Online survey of 9-16 year olds and their parents in 25 countries. Retrieved from London School of Economics and Poli cal Science website: http://eprints.lse.ac.uk/33731/
Livingstone, S., Mascheroni, G., & Staksrud, E. (2015). Developing a framework for researching children’s online risks and opportuni es in Europe. Retrieved from London School of Economics and Poli cal Science website: http://eprints.lse.ac.uk/64470/1/__ lse.ac.uk_storage_LIBRARY_Secondary_libfile_shared_repository_Content_EU%20Kids%20Online_EU%20Kids%20Online_ Developing%20framework%20for%20researching_2015.pdf
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 19 Livingstone, S., & Mason, J. (2015). Sexual rights and sexual risks among youth online: A review of exis ng knowledge regarding children and young people’s developing sexuality in rela on to new media environments. Retrieved from London School of Economics and Poli cal Science website: http://eprints.lse.ac.uk/64567/1/Livingstone_Review_on_Sexual_rights_and_sexual_risks_among_ online_youth_Author_2015.pdf
Madden, M., Cortesi, S., Gasser, U., Lenhart, A., & Duggan, M. (2012). Parents, teens, and online privacy. Retrieved from Pew Research Center website: https://www.pewinternet.org/2012/11/20/parents-teens-and-online-privacy/
Madigan, S., Villani, V., Azzopardi, C., Laut, D., Smith, T., Temple, J. R., ... Dimitropoulos, G. (2018). The prevalence of unwanted online sexual exposure and solicita on among youth: A meta-analysis. Journal of Adolescent Health, 63(2), 133–141. https: //dx.doi.org/10.1016/j.jadohealth.2018.03.012
Mar n, F., Wang, C., Pe y, T., Wang, W., & Wilkins, P. (2018). Middle school students’ social media use. Journal of Educa onal Technology & Society, 21(1), 213–224. Retrieved from https://www.j-ets.net/ETS/journals/21_1/19.pdf
Mascheroni, G., & Cuman, A. (2014). Net children go mobile: Final report (with country fact sheets). Deliverables D6.4 & D5.2. Retrieved from Net Children Go Mobile website: http://netchildrengomobile.eu/reports/
McKenna, K., Green, A., & Gleason, M. (2002). Rela onship forma on on the internet: What’s the big a rac on? Journal of Social Issues, 58(1), 9–31. https://dx.doi.org/10.1111/1540-4560.00246
Menn, J. (2018, October 15). Exclusive: Facebook to ban misinforma on on vo ng in upcoming U.S. elec ons. Retrieved from https: //www.reuters.com
Meyer, M., Adkins, V., Yuan, N., Weeks, H. M., Chang, Y., & Radesky, J. (2019). Adver sing in young children’s apps: A content analysis. Journal of Developmental & Behavioral Pediatrics, 40(1), 32–39. Retrieved from http://childrenstech.com/files/2018/ 11/Advertising_in_Young_Children_s_Apps___A_Content.99257.pdf
Mitchell, K., Ybarra, M., & Korchmaros, J. (2014). Sexual harassment among adolescents of different sexual orienta ons and gender iden es. Child Abuse and Neglect, 38(1), 43–71. https://dx.doi.org/10.1016/j.chiabu.2013.09.008
Nestler, E. J., & Malenka, R. C. (2004). The addicted brain. Scien fic American, 290(3), 78–85, Retrieved from http://sites.oxy.edu/ clint/physio/article/theaddictedbrain_sciam.pdf
Nixon, C. L. (2014). Current perspec ves: The impact of cyberbullying on adolescent health. Adolescent Health, Medicine and Ther- apeu cs, 5, 143–158.
Nyst, Carly. (2017). Privacy, protec on of personal informa on and reputa on. Retrieved from Unicef website: https://www.unicef. org/csr/css/UNICEF_CRB_Digital_World_Series_PRIVACY.pdf
Ohlheiser, A. (2013, December 13). There are 13 countries where atheism is punishable by death. The Atlan c. Retrieved from https://www.theatlantic.com
Palfrey, J. G., & Gasser, U., & boyd, d. (2010). Response to FCC no ce of inquiry 09-94: Empowering parents and protec ng children in an evolving media landscape (Berkman Center Research Publica on No. 2010-02). Retrieved from https://cyber.harvard.edu/ publications/2010/Re_Empowering_Parents_Protecting_Children
Phyfer, J., Burton, P., & Leoschut, L. (2016). South African kids online: Barriers, opportuni es & risks. A glimpse into South African children’s internet use and online ac vi es. Retrieved from Global Kids Online website: http://globalkidsonline.net/wp-content/ uploads/2016/06/GKO_Country-Report_South-Africa_CJCP_upload.pdf
Privacy Program. (2019). Retrieved from https://privacy.commonsense.org
Quraishi, H. (2019, April 13). Under employers’ gaze, gen Z is bi ng its tongue on social media. Retrieved from https://www.npr.org
Ringrose, J., Harvey, L., Gill, R., & Livingstone, S. (2013). Teen girls, sexual double standards and ‘sex ng’: Gendered value in digital image exchange. Feminist Theory, 14(3), 305–323. https://dx.doi.org/10.1177/1464700113499853
Schmidt, S. (2017, August 2). A er months of bullying, her parents say, a 12-year-old New Jersey girl killed herself. They blame the school. The Washington Post. Retrieved from https://www.washingtonpost.com
Sengupta, S. (2013, October 13). Privacy fears grow as ci es increase surveillance. The New York Times. Retrieved from https: //www.nytimes.com
20 PRIVACY RISKS AND HARMS privacy.commonsense.org Shaw, J. (2017, January/February). The watchers: Assaults on privacy in America. Harvard Magazine. Retrieved from https: //harvardmagazine.com
Sherman, L., Payton, A., Hernandez, L., Greenfield, P., & Dapre o, M. (2016). The power of the like in adolescence. Psychological Science, 27(7), 1027–1035. https://dx.doi.org/10.1177/0956797616645673
Singer, N. (2018, July 26). Amazon’s facial recogni on wrongly iden fies 28 lawmakers, A.C.L.U. says. The New York Times. Retrieved from https://www.nytimes.com
Singer, P. (2018, August 29). Kids 35 mes more likely to be iden ty the vic ms. Here’s how to avoid it. The Democrat & Chronicle. Retrieved from https://www.democratandchronicle.com
Slavtcheva-Petkova, V., Nash, V., & Bulger, M. (2015). Evidence on the extent of harms experienced by children as a result of online risks: Implica ons for policy and research. Informa on, Communica on & Society, 18(1), 48–62. https://dx.doi.org/10.1080/ 1369118X.2014.934387
Stucker, M. (2014, March 2). Girl costs father 80,000 dollars with ‘SUCK IT’ Facebook post. Retrieved from http://www.cnn.com
Talbot, D. (2012, September 12). How Facebook drove voters to the polls. MIT Technology Review. Retrieved from https://www. technologyreview.com
Third, A., Bellerose, D., Dawkins, U., Kel e, E., & Pihl, K. (2014). Children’s rights in the digital age: A download from children around the world. Retrieved from Western Sydney University website: https://www.uws.edu.au/__data/assets/pdf_file/0003/753447/ Childrens-rights-in-the-digital-age.pdf
Tiku, N. (2018, September 18). ACLU says Facebook ads let employers favor men over women. Wired. Retrieved from https: //www.wired.com
Tokunaga, R. (2010). Following you home from school: A cri cal review and synthesis of research on cyberbullying vic miza on. Computers in Human Behavior, 26, 277–287. https://dx.doi.org/10.1016/j.chb.2009.11.014
Unicef. (2017). The state of the world’s children 2017: Children in a digital world. Retrieved from https://www.unicef.org/publications/ files/SOWC_2017_ENG_WEB.pdf
Winters, K. C. (2009). Adolescent brain development and alcohol abuse. The Journal of Global Drug Policy and Prac ce, 3(3). Retrieved from https://www.dfaf.org/wp-content/uploads/2018/11/Vol-3-Issue-3.pdf
Winters, K. C., & Arria, A. (2011). Adolescent brain development and drugs. The Preven on Researcher, 18(2), 21–24. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3399589/
Wri en statement to the United States Senate Commi ee on the Judiciary, in the ma er of Cambridge Analy ca and other related issues, 115th Cong. 22 (2018) (tes mony of Christopher Wylie).
Zuboff, S. (2015). Big other: Surveillance capitalism and the prospects of an informa on civiliza on. Journal of Informa on Technology, 30, 75–89. https://dx.doi.org/10.1057/jit.2015.5
Zuboff, S. (2016, May 3). Wie wir Googles Sklaven wurden. [The secrets of surveillance capitalism]. Frankfurter Allgemeine Zeitung. Retrieved from https://publicpurpose.com.au/wp-content/uploads/2016/04/Surveillance-capitalism-Shuboff-March-2016. pdf
CREATIVE COMMONS ATTRIBUTION 4.0 INTERNATIONAL PUBLIC LICENSE PRIVACY RISKS AND HARMS 21 OUR OFFICES
San Francisco Headquarters 650 Townsend Street, Suite 435 San Francisco, CA 94103 (415) 863-0600
Los Angeles Office 1100 Glendon Avenue, 17th Floor Los Angeles, CA 90024 (310) 689-7535
New York Office 575 Madison Avenue New York, NY 10022 (212) 315-2138
Washington, D.C. Office 2200 Pennsylvania Avenue NW 4th Floor East Washington, D.C. 20037 (202) 350-9992
www.commonsense.org
© 2019 Common Sense Media. All rights reserved. Common Sense, associated names, associated trademarks, and logos are trademarks of Common Sense Media, a 501(c)(3) nonprofit organiza on, FEIN 41-2024986.