Privacy, safety, and the role of parents: Gaps in user understanding of risk and solutions and how we can support users to make better privacy decisions

Raw text

Transcribed text as recorded, including headers and formatting fragments.

CONFIDENTIAL CONFIDENTIAL Document Provided in Native Format PLTF EX PLTF EX-01350 ΜΕΤΑ3047MDL-047-00346070 METANMAG-005-01031694 PLTF EX-01350_0001 Privacy, safety, and the role of parents Gaps in user understanding of risk and solutions and how we can support users to make better privacy decisions Hitomi Hayashi- Branson Product Researcher Internal only - Not for distribution Slide 1 Notes Team/Research/XFN presentations (2 options): TBD L&L: 8/25 Overview 1. Setting the stage - A quick overview of what we know about teen safety 2. Factors that prevent teens and parents from taking action a. They're in the dark when it comes to risks and solutions b. They don't know about the tools we have or just don't want to use them 3. Privacy - Attitudes and approaches a. Privacy is not top of mind and viewed as limiting b. The role of parents/guardians 4. Adapting privacy to users' needs - A proposal for how privacy can evolve alongside age, needs, capabilities, and maturity. TL;DR: Key findings (especially those in settings), they don't know how they work, they use them reactively, or they just don't want to use them. 2. Teens don't want to shut the door to anyone. Many teen don't want to say “no”, aren't accustomed to asking questions, fear missing out (on opportunities or connections), or don't want to seem rude. This drastically elevates their risk. 3. Teens and parents have limited knowledge of the risks online and don't know what to do when encountering bad actors. Teens and parents start their Instagram journeys with little warning and preparation for what can happen to them online. Parents have varying level of digital literacy and a general low level of social media experience. Many parents feel lost and confused on how they can protect their children. 4. Parents' desired and actual level of involvement varies greatly (from not involved to very involved) and evolves as the child matures. Generally, teens are open to parents playing a larger role when they're younger (13-14 yos) but expect to be more independent as they grow and show signs of maturity. 5. Users need help managing their safety and privacy online over time. While most teens have limited knowledge of the tools that exist, most want more control over their Instagram experience in a way that supports their needs to connect and interact with others (including new people) in a safe way. Opportunities tney start instagram, contextually (e.g., wnen naving a pad experience) and periodically (by making education easily accessible). Explore ways to communicate and demonstrate the impact of using/turning on specific tools and controls (e.g., via tools, dashboards, and feedback mechanisms). 2. Empower teens. Educate/show teens how to have a safer (and better experience) online (e.g., by keeping their circle smaller), provide them with strategies for assessing risk, and empower them to question others and to say “no”. Explore ways to devalue random follows and request (e.g., highlighting low quality or negative interactions from these people). 3. Educate and raise awareness of risks online and the tools we have. Develop education for a range of user and supporter types to raise awareness of risks. See opportunity 1 on tool education. 4. Provide parents/guardians and supporters with a range of low to high effort/involvement options to support teens. Provide parents/guardians with different options and level of control based on theirs' and their child's desired involvement. Explore ways to facilitate decreased (or increased) control over time and events. 5. Promote the use of privacy tools by making it easier to use them. In addition to upsells and education, suggest settings that support users across different life-stages, maturity, and different use case (e.g., “Help me expand my network in a safer way”, “I'm looking to connect with teens like me"). Make these easier to turn on and empower users with education on how they work as well as the ability to easily adjust settings to match their needs. 3 Internal only - Not for distribution Slide 3 Notes Interviews with external child safety partner organizations Setting the stage A quick overview of what we know about teen safety. Internal only - Not for distribution What we know What do we know so far about inappropriate content and interactions with children on IG? It's common and expected According to market SMEs, external partners, teens and parents, seeing inappropriate images of children and having inappropriate interactions are common and even expected, especially among young girls. And it takes many forms: Beyond CEI, IIC, & MS, these include (but not limited to): • • • Sharing CEI CEI solicitation NCII Sextortion And more! And it's not always obvious! Consequences OKOSOVOKO Victimization can lead to physical health, mental health, and social consequences on and offline...and can cause users to feel unsafe to share & express themselves (see research here and here). Larger graphs and (possibly) uncertainty in who can see and interact with you can lead to low production of content. Lastly, in data analysis on bullying, we saw how having negative experiences make users more likely to delete or deactivate their account. We should seek to understand if this finding is also true for teen safety threats. For details on the above points, refer to this summary of past research. Actors range in type and tactics There are a wide range of types. While, according to experts they are mostly men, they can be: women, other teens, fake or real accounts, domestic or foreign actors etc. Actors are comfortable on Instagram and use a range of surfaces and tools on IG to view inappropriate content, network, share and sell content with seemingly no consequences. Actors are extraordinary savvy and know just how to disarm & develop relationships with kids. Targets range in type and risk Common targets (which vary by country) are: • • • • Women People of color LGBTQ+ Other marginalized groups Creators/large public accounts Users who are struggling financially or in other ways Additionally, high risk moments (e.g., COVID) can causes elevated risk due to more time spent online, loneliness, and stress. Moving forward, we should seek to be more proactive identifying and intervening on these moments Factors that prevent teens and parents from taking action - They're in the dark when it comes to risks and solutions - They don't know about the tools we have or just don't want to use them Internal only - Not for distribution ດ Users don't have a full understanding of risks... Users have limited understanding about what happens online and what to do when it happens. Teens start Instagram with little knowledge of the dangers that exist. Some issues and observations. - - The new user experience (NUX) lacks deep education on risks and safety on Instagram Many don't know what to expect (especially with a public profile)* Most learn about the dangers online in real time and on their own Most don't know about the tools that we have** Most use tools reactively “I started off with a private account because my mom wanted me to. After 6 months I changed to a public account. I was nervous when I first did this because I'd heard stories of the creepy people on the internet.” Woman, 14 years old *For related research on privacy and expectations, see this study. **For related research on awareness of tools, see this survey. Internal only - Not for distribution Slide 7 Notes • And while self reliance and seeking support for peers and parents are common, most of these people have very limited knowledge. Ο Parents have limited knowledge on what's going on and don't know where to start. Many parents lack an understanding of the real risks that exist online. In combination with varying level of digital literacy and a general low level of social media experience, many parents feel lost and confused on how they can protect their children. ດ ...and neither do parents/guardians Parents/guardians have limited understanding about what happens online and what to do when it happens. According to experts, parents/guardians have limited knowledge on what's going on and don't know where to start. Many parents/guardians lack an understanding of the real risks that exist online. In combination with varying level of digital literacy and a general low level of social media experience, many parents/guardians feel lost, overwhelmed, and hopeless. "Parents have to learn how to keep their kids safe on Instagram, then there's Youtube, and Twitter. It's just so stressful that parents will just throw their hands up - give up." External partner "The education that you have today isn't 'real' enough. You have to tell people about the dangers and what exactly their kids can experience online. People need to know.” External partner - *For related research on privacy and expectations, see this study. **For related research on awareness of tools, see this survey. Internal only - Not for distribution Slide 8 Notes • And while self reliance and seeking support for peers and parents are common, most of these people have very limited knowledge. Ο Parents have limited knowledge on what's going on and don't know where to start. Many parents lack an understanding of the real risks that exist online. In combination with varying level of digital literacy and a general low level of social media experience, many parents feel lost and confused on how they can protect their children. Parents/Guardians need a lot of help Experts and users emphasized how lost and in the dark parents are regarding the dangers and risks online. ດ They might not know what “suspicious” looks like... “A man messaged me and asked me to be a brand ambassador for his lingerie company. He asked for photos of me in my underwear. I showed my mom and we were both skeptical but we decided to investigate to see if it was a real opportunity. We found out that the account was fake, so I stopped talking to him at that point.” Teen, woman They might not realize the risks... "My dad is actually the one who told me I should have a public account so more people could see my photography. He didn't really know about public accounts [and the risks], but I had heard stories from friends so I was a bit nervous.” Teen, woman Some promote their child not thinking of the consequences... "These Instagram moms post pictures of their kids in bikinis and then are surprised when they get inappropriate comments from men.” External partner Internal only - Not for distribution Slide 9 Notes Beyond demographics and life events... • Who is at risk Ο Ο Ο Ο Ο Ο Children don't want to shut the door to any interactions putting them at high risk. Some are driven by the need for followers while others seek attention. This makes them vulnerable to bad actors who understand these needs. Women, LGBTQ+, and financially struggling users are common targets across the board. However, risk groups differed by country. For example, those in financial need were most commonly mentioned by Persian, Portugese, Romanian, Sub-saharan Africa, and Thai market SMEs. Young women are the most commonly victimized group. They are also most likely to produce/distribute/sell their own content for the purpose of earning money or due to mental or emotional health issues (e.g., low self-esteem). In ESLA, cultural norms that are permissive of older men dating younger women also facilitates these interactions. Financial need elevates risk of victimization by 1) making it more likely for users to produce and sell their own content/bodies and 2) giving actors an easy way to draw in victims (i.e., by offering them money or opportunities). On this second point, groomers in particular can easily draw in unsuspecting and trusting victims with promises thereby gaining the trust of these users. American and Indian girls experience more solicitation and inappropriate interactions with adults: e.g. adults asking for personal information, nude photos, threatening with blackmail, etc. American boys' experiences were generally limited to uncomfortable interactions whereas Indian boys experienced interactions that were more intense, though they expressed more shame in discussing their experiences compared to American boys. Black American teens experienced the most intense interactions, among boys and girls. Opportunities: Slide 9 Notes (Continued) Implement surface-specific interventions and education based of common surfaces by country. Partner with other relevant apps (e.g., Line in Japan) to stop actors from moving between apps. Educate users on likely actors and strategies based on what we see in each specific country. Provide additional support, outreach, and education to particularly high-risk users - i.e., women, children in poor or rural communities, LGBTQ youth etc. Users don't take advantage of the tools we have Users are unaware of many tools*... In a survey measuring users' awareness of tools, while most users (>50%) were aware of contextual tools like block and delete, a significantly smaller proportion of people were aware of tools in settings like comment filter and mention controls. ...don't know what tools do In the above survey, while most users rated their own knowledge of tools to be at least fair, in reality, a large proportion of users did not know (or were unsure) about how tools worked. While some misunderstanding are inconsequential, others have major consequences (e.g., leaving users exposed to harm). ...use tools reactively Based on research with teens and adults in the US and India, users typically don't use proactively or explore tools until they (or someone they know) has a negative experience. ...or just don't want to use tools In past research with regular users and creators, people might not use tools due to concerns about limiting positive interactions or opportunities. “Even if I get [inappropriate requests] I don't block DMs. What if it's an opportunity or someone that I met at a party? I'd rather just have the DMs there.” - Teen, woman *Findings are from this survey on awareness of tools among Instagram users. See related research on awareness of tools on messenger here (POC: Meenakshi Mennon). And FB privacy research here (POC: Denise Sauerteig and Sarah Kling). Internal only - Not for distribution Slide 10 Notes • What do users do? Maybe rely on themselves Ο Abuse is mitigated through self-reliance and varies by age. In both the U.S. and India, teens were largely self-reliant and had developed their own processes for assessing risk and mitigating abusive behavior online. However, self-reliance differed by age: i. ii. iii. 13-14: Younger teens in both markets more often engaged interactions to avoid being rude to an adult, and often relied on their parents for guidance on what to do in uncomfortable situations. 15-17: Older teens in both markets had grown used to blocking unwanted interactions immediately and not disclosing to their parents unless their physical safety was being threatened. Teens in both countries of all ages also strongly relied on their friends and communities to mass report bad actors. Teen hesitate to limit interactions and to say "no" (open) This puts teens at increased risk of unwanted and inappropriate interactions. Some specific behaviors and attitudes that endanger teens include: • • Desire for followers (e.g., growing creators) Desire for friends Desires for attention Not wanting to feel limited or closed off from the world Not being accustomed to questioning others' intent/Being very trusting Fear of missing out on opportunities and relationships Having a hard time saying "No” or rejecting others Not wanting to seem rude Internal only - Not for distribution Slide 11 Notes Beyond demographics and life events... • Who is at risk Ο Ο Ο Ο Ο Children don't want to shut the door to any interactions putting them at high risk. Some are driven by the need for followers while others seek attention. This makes them vulnerable to bad actors who understand these needs. Women, LGBTQ+, and financially struggling users are common targets across the board. However, risk groups differed by country. For example, those in financial need were most commonly mentioned by Persian, Portugese, Romanian, Sub-saharan Africa, and Thai market SMEs. Young women are the most commonly victimized group. They are also most likely to produce/distribute/sell their own content for the purpose of earning money or due to mental or emotional health issues (e.g., low self-esteem). In ESLA, cultural norms that are permissive of older men dating younger women also facilitates these interactions. Financial need elevates risk of victimization by 1) making it more likely for users to produce and sell their own content/bodies and 2) giving actors an easy way to draw in victims (i.e., by offering them money or opportunities). On this second point, groomers in particular can easily draw in unsuspecting and trusting victims with promises thereby gaining the trust of these users. American and Indian girls experience more solicitation and inappropriate interactions with adults: e.g. adults asking for personal information, nude photos, threatening with blackmail, etc. American boys' experiences were generally limited to uncomfortable interactions whereas Indian boys experienced interactions that were more intense, though they expressed more shame in discussing their experiences compared to American boys. Black American teens experienced the most intense interactions, among boys and girls. Opportunities: Slide 11 Notes (Continued) Implement surface-specific interventions and education based of common surfaces by country. Partner with other relevant apps (e.g., Line in Japan) to stop actors from moving between apps. Educate users on likely actors and strategies based on what we see in each specific country. Provide additional support, outreach, and education to particularly high-risk users - i.e., women, children in poor or rural communities, LGBTQ youth etc. Privacy - Attitudes and approaches - - Privacy is not top of mind and viewed as limiting The role of parents/guardians Internal only - Not for distribution Privacy is NOT top of mind We've established in past research (e.g., here, here) that privacy and safety are not top of mind for users. Experts wanted us to find a way to bring these issue to the forefront in a way that accurately represents risks (without causing unnecessary alarm). Also see the above slide on why users don't use the tools we have. “I never really explored the settings until now. I really wish I had known about these. If I had know, I wouldn't have deleted my account when I was being harassed." Woman, teen "Privacy and safety aren't things that teens will think about when they start Instagram. They just want to use Instagram and start posting. We have to find a way to communicate risks alongside the benefits of using Instagram.” External partner Internal only - Not for distribution More private is seen as more limited Private accounts and turning on controls are typically seen as limiting and not conducive to those who want to grow and connect with new people. We also heard this from past research with users and creators. Private = Safe, Public = Seen In research with teens and adults, we heard the a private account is associated with "staying safe” whereas a public account is associated with "being seen". Note that public accounts were not associated with being "unsafe”. Instead, public account users accepted some level of risk and expected to have to implement strategies to stay safe online (e.g., screening messages, being careful about what they posted, not sharing location). LIMITS “I used to switch back and forth from private to public a lot. My mom was over protective and wanted me to stay safe.” “Ultimately, I wanted my presence out there, which is why I didn't switch back to a private account." Woman, 18 years old [After going through setting for the first time] “This is better than I thought... if you're public, you can have a lot more privacy than people realize...” Woman, 13 years old How can we create a public experience that is also associated with being safe? Internal only - Not for distribution Parents/Guardians play a role in users' privacy which typically evolves as the child mature Generally, younger teens lean heavily on support from parents whereas older teens are more independent. Note that the below categories are suggestions established from user interviews. 13-14: "Learning the ropes" More often engage in interactions to avoid being rude to an adult, and often relied on their parents for guidance on what to do in hard situations. Parents might have rules in place to keep their child safe (e.g., starting with a private account, not talking to strangers, visibility into DMs). 15: "Gaining independence" Start to develop processes to deal with uncomfortable situations and Increased independence from parents based on proven abilities. *See DS work by Aashna Pandya (e.g., here, here) for more on difference across life- stages. Reach out to Jessica Weber for upcoming work on this topic. 16-17: "Becoming independent” Largely self-reliant and had developed their own processes for assessing risk and mitigating abusive behavior online. Used to blocking unwanted interactions immediately and not disclosing to their parents unless their physical safety was being threatened. Internal only - Not for distribution Slide 15 Notes • What do users do? Maybe rely on themselves Ο Abuse is mitigated through self-reliance and varies by age. In both the U.S. and India, teens were largely self-reliant and had developed their own processes for assessing risk and mitigating abusive behavior online. However, self-reliance differed by age: i. ii. iii. 13-14: Younger teens in both markets more often engaged interactions to avoid being rude to an adult, and often relied on their parents for guidance on what to do in uncomfortable situations. 15-17: Older teens in both markets had grown used to blocking unwanted interactions immediately and not disclosing to their parents unless their physical safety was being threatened. Teens in both countries of all ages also strongly relied on their friends and communities to mass report bad actors. ... which is beyond just age* In research with teens and external partners, some signs of maturity include: • Showing you can manage difficult situations Blocking and reporting suspicious people Not interacting with strangers Ο Ο Not sharing personal information • • • Posting mature and healthy content. Interacting with others in healthy ways. Viewing age appropriate content. "There is no differences between a 12 and 13 year old. It's more about their personal maturity." - Man, 16 years old “My parents trusted us from a very young age because they knew I was mature enough and that I would come to them if I needed help. - Woman, 14 years old *See related findings on what maturity might look like here (POC: Laura Brown). Internal only - Not for distribution Parents/Guardians vary greatly in how involve they ar Parenting style, capabilities, needs, and other factors also determine parental involvement. Parents' desired and actual level of involvement varies from not involved to very involved.* How involved do parents want to be? In research with parents in the U.S. and India, we found that there is a wide range in parents' needs, practices, and preference for control when it comes to keeping their kids safe online. In a study we saw country-level differences: • American parents did not have a desire for more control, but are open to more visibility • Indian parents want both more visibility and control. How involved do teens want their parents to be? • Users have different current and desired levels of parental involvement based on their age and maturity. Users expected parental involvement to decrease as they grew and matured. • • • In both countries, teens hesitate to give their parents more control but are generally open to giving them more visibility on discrete elements of their social media** (e.g. Parents can see DMs from strangers but not from friends). Note that in past research (here, here) teens also search for support outside of parents. *See related research on relationship within the household shape users' experience online (POC: Kenzie Snyder), and research on to role of bystanders (POC: Millie Harrison). **See related suggestions from research with experts on giving parents more control and insights. See recent research on parental involvement on Whatsapp here (POC: Marta Ponte Fissgus). Internal only - Not for distribution Slide 17 Notes • Parental involvement varies greatly within nad across countries. They play varying roles and have different levels of knoweldge. Ο Ο they lean on others for suppot (and parent lay a key role especaily for younger uesrs) i. Parental support is an important part of teens' online safety. In both the U.S. and India, parents have different needs and practices when it comes to keeping their kids safe online. 1. American parents do not have an active desire for more control, but are open to more visibility while parents in India want both more visibility and control. 2. In both countries, teens hesitate to give their parents more control but are open to giving them more visibility on discrete elements of their social media (e.g. DMs from strangers are okay to share, but not DMs from friends/family members). Participants had different current and desired levels of parental involvement based on their age and maturity. Note that these participants were generally close with parents and kept them (at least peripherally) involved with their social media use. Adapting privacy to users' needs A proposal for how privacy can evolve alongside age, needs, capabilities, and maturity. Internal only - Not for distribution We have public & private...what about something in between? We heard from participants that a public account can be too open. Users wanted a middle ground (between private and public) that would allow them to expand their reach/network while keeping them safe and in control. What could this look like? Recommended settings packages based on different use cases (e.g., “Help me expand my network in a safer way”, “I'm looking to connect with teens like me”)*. The example to the right, a “Semi-Public" account, is a public account that allows users to narrow in on their network (including or excluding certain groups). Given users' desire for control, this should be accompanied by explanation, education, and (possibly) the option to adjust settings. For more on how we can balance simplicity, complexity, transparency, and control, see this study. Existing controls Interactions (Comments, DM, Tags, Mentions) New controls Follow request requirement Visibility (to others) Discoverability (e.g., account or content suggested or promoted) Searchability "Semi-Public"/Something in between Option to limit interaction to only include certain groups (e.g., only teens, people in my area), or exclude certain groups (e.g., unconnected users, non-teens, suspicious accounts: fake accounts, bots, bad actors) Required for certain groups of people Followers and non-followers excluding certain groups of people Excluding certain groups of people Excluding certain groups of people *See related research on a number of recommended settings packages based on user needs here. Also see default settings recommendations from the Privacy team here (POC: Mary Titus). Internal only - Not for distribution The evolution of privacy + safety over time For a more detailed version of the below table, see this quip. Also see the Privacy team's recommendations & plan for defaulting ad hoc privacy settings to: 1) align setting with user expectations and 2) default private teens into safer visibility, comment, tag, mention, and DM controls (POCs: Mary Titus, Kathie Pham). Interactions Account type (Comments, (a) DM, Tags, Mentions) Age group ~10-12 yo Extra-private Start Private, (b) -13-15 yo "Semi-public" Followers only Private - Defaulted to followers only "Semi- public" - Followers + option to expand to Content Discoverability (e.g., account or content suggested Searchability they see (c, d) or promoted) Follow request requirement Visibility to others Required for all Followers only Tweens only Parental Child's level of control Restricted from No Education involvement requirement (e) (f) Tweens only adult content + (independent) Required filter control Required Security requirement 2-fac required, linked to parent account Private - Required for all Private - Restricted from Followers only adult content "Semi-public" "Semi-public" Teens only - Followers Teens only (with option for parental approval) + filter Partial control Suggested with Strongly 2-fac required suggested - Required for non-teens and and teen non- suspicious accounts. followers opt out teen non- followers Less Private, "Semi- -16-17 yo public", Public Dependent on account choice and settings (a) For privacy needs/expectations across lifestage, work by Aashna Pandya (e.g., here, here). restricted/Unr Complete estricted + filter control Low Suggested 2-fac required (b) A semi-public account would allow users to expand their reach and network while keeping them safe and in control of who can and can not see and interact with them. For details on how a “Semi-public” account could work, refer to this quip. (c) Some wanted the ability to block out certain types of content or content from certain (types of) accounts. (d) In a past study on age verification here teens said that being blocked from seeing certain content would be annoying but, for the most part, understandable. (e) We saw in past research with teens in the US and India (here and here) that parental involvement and privacy expectations differed greatly by culture. (f) Past research with external experts highlighted the importance of education aimed at preparing users for what they might face on Instagram. Internal only - Not for distribution Summary: Key findings (especially those in settings), they don't know how they work, they use them reactively, or they just don't want to use them. 2. Teens don't want to shut the door to anyone. Many teen don't want to say “no”, aren't accustomed to asking questions, fear missing out (on opportunities or connections), or don't want to seem rude. This drastically elevates their risk. 3. Teens and parents have limited knowledge of the risks online and don't know what to do when encountering bad actors. Teens and parents start their Instagram journeys with little warning and preparation for what can happen to them online. Parents have varying level of digital literacy and a general low level of social media experience. Many parents feel lost and confused on how they can protect their children. 4. Parents' desired and actual level of involvement varies greatly (from not involved to very involved) and evolves as the child matures. Generally, teens are open to parents playing a larger role when they're younger (13-14 yos) but expect to be more independent as they grow and show signs of maturity. 5. Users need help managing their safety and privacy online over time. While most teens have limited knowledge of the tools that exist, most want more control over their Instagram experience in a way that supports their needs to connect and interact with others (including new people) in a safe way. Opportunities they start instagram, contextually (e.g., wnen naving a pad experience) and periodically (by making education easily accessible). Explore ways to communicate and demonstrate the impact of using/turning on specific tools and controls (e.g., via tools, dashboards, and feedback mechanisms). 2. Empower teens. Educate/show teens how to have a safer (and better experience) online (e.g., by keeping their circle smaller), provide them with strategies for assessing risk, and empower them to question others and to say “no”. Explore ways to devalue random follows and request (e.g., highlighting low quality or negative interactions from these people). 3. Educate and raise awareness of risks online and the tools we have. Develop education for a range of user and supporter types to raise awareness of risks. See opportunity 1 on tool education. 4. Provide parents/guardians and supporters with a range of low to high effort/involvement options to support teens. Provide parents/guardians with different options and level of control based on theirs' and their child's desired involvement. Explore ways to facilitate decreased (or increased) control over time and events. 5. Promote the use of privacy tools by making it easier to use them. In addition to upsells and education, suggest settings that support users across different life-stages, maturity, and different use case (e.g., “Help me expand my network in a safer way”, “I'm looking to connect with teens like me"). Make these easier to turn on and empower users with education on how they work as well as the ability to easily adjust settings to match their needs. 21 Internal only - Not for distribution Slide 21 Notes Interviews with external child safety partner organizations Safe environmen t Stop bad actors Support and empower targets Educate • Completed and Ongoing work Detect potential violators Continue to improve detection and removal of content and actors Planned work* • Tackling minor sexualization • Hashtag and search education • Making it more difficult for adults to find/follow teens (Update: here) Restricting DMs between teens and adults Stop recommending minor accounts in suggestions Restrict discovery of minor content on Reels tab • Prompting teens to be more cautious about DMs • Prevent CEI • Age verification Understand/Prepare for E2EE • Breaking up networks Making it harder for bad actors to find content : Protecting minors Understand at-risk groups and opportunities • Privacy and • Safety New parents guide Encouraging teens to make their accounts private Defaulting teens to private *Research roadmap; Team roadmap • Tween, teen, parent/guardian education • Safety defaults Privacy settings/packages Slide 23 Notes *Research plan: https://fb.quip.com/CpjEA1easSP8#bAZACAToacH; Team roadmap: https://fb.quip.com/F3yQApxf73Q7