The Screenagers Family AI Agreement, Section by Section

A family AI agreement is a short written plan that says how your household uses AI chatbots, image tools and AI features in apps, and why. The one most parents are seeing this fall comes from the team behind the Screenagers films, released as a free printable on October 27, 2026. It has six sections and a pledge for parents. It is a good template, but it treats “kids and teens” as one group. This article goes through it section by section, checks each one against the research, and shows how to rewrite it for each age. Spoiler: the section with the best evidence behind it is the one about sleep, and the most important one is the part parents sign.

Key Takeaways

• The Screenagers template has six sections (social life, academic life, mental wellness, sleep, fakes, privacy), each with a guiding question, examples to discuss and blank space for your own rules.
• Its final “Safety-First Commitment” asks parents to promise a team response instead of reflexively taking away tech. Research on teen disclosure backs this up.
• The sleep section rests on the strongest evidence: a meta-analysis of 125,198 children linked bedtime device access, even without use, to worse sleep.
• The academic section is well supported by a 2025 field experiment: students who used unrestricted ChatGPT for practice scored 17% worse once it was taken away.
• For a 7-year-old, turn each section into one concrete house rule. For a 14-year-old, let them draft the rules and negotiate the why.

What the Screenagers Family AI Agreement Is

The Screenagers Family AI Agreement is a fill-in template titled “Family AI Agreement For Kids & Teens: A Roadmap for Supporting Healthy Development.” Each of its six areas opens with a question and lists examples to talk through. There are no pre-written rules. Families write their own “agreed-upon guidelines,” sign it and pick a date to review it. The page that hosts it says the goal “isn’t perfect rule-following but ongoing conversations about our WHYs,” and recommends revisiting it every few months.

It fits inside a bigger frame. In January 2026, the American Academy of Pediatrics replaced its media guidance with a policy statement called “Digital Ecosystems, Children, and Adolescents.” Its family advice: make a Family Media Plan, model healthy use, set media-free times and places, and use its “5 Cs” (Child, Content, Calm, Crowd-out, Communication). The Screenagers template is, in effect, an AI chapter for that plan. If you already have screen rules, see what the research says about family screen time agreements that actually work.

Now, section by section.

Section 1: Social Life

The template asks: “What guidelines should there be for interacting with AI chatbots that communicate in human-like ways?” Examples include going to AI for advice, sharing feelings with AI companions, chatting with AI instead of friends or family, turning to AI when lonely and keeping AI conversations secret.

What the research says. Use is common. Common Sense Media’s July 2025 report on AI companions found nearly three in four teens had used them, half used them regularly, and a third had chosen an AI companion over a human for a serious conversation. A small September 2026 preprint by Jiaying Liu and Nimra Ishfaq (17 young adults, not peer-reviewed) found participants worried chatbots could discourage reciprocity, the back-and-forth of supporting each other. They proposed AI that supplies “materials to engage with others”: help figuring out what to say to a real person.

Verdict: well grounded on prevalence; effects are still being studied. The “secret conversations” example is the one to prioritize, because secrecy cuts you off from everything else.

For a 7-year-old: “Talking toys and chatbots are not friends. If one says something weird or asks you to keep a secret, tell a grown-up.”

For a 14-year-old: “If you use AI to vent or practice a hard conversation, that’s OK. If it’s starting to replace talking to friends, let’s notice that together.” Ask them to suggest the line.

Section 2: Academic Life

The template asks: “What guidelines should there be for interacting with AI chatbots regarding schoolwork?” Examples: following teachers’ rules, brainstorming with AI or not, asking AI to solve problems versus trying first, submitting AI work as one’s own, using AI to explain concepts and to make practice quizzes.

What the research says. This section has unusually good evidence. Hamsa Bastani and colleagues ran a field experiment with nearly 1,000 high school math students, published in PNAS in 2025. Students with a standard ChatGPT-style tutor did 48% better on practice problems while they had it. When access was removed, they scored 17% worse than students who never had it. A second version, prompted to give hints instead of answers, largely avoided that drop. The authors’ word for the first pattern was “crutch.”

Verdict: strongly supported. The template’s distinction between “solve it for me” and “explain it to me” is exactly the line the experiment drew.

For a 7-year-old: Mostly not relevant yet. One rule: “Ask me or your teacher before using AI for homework.”

For a 14-year-old: “Try first, then ask AI to explain, never to do. Follow each teacher’s rule.” Have them name which of their classes allow AI and how.

Section 3: Mental Wellness

The template asks: “What guidelines should there be for using AI for life advice or emotional support?” Examples: relationship advice or big decisions, treating AI like a therapist or best friend, turning to AI instead of a person during distress, and knowing whom to contact in a mental health crisis.

What the research says. A Common Sense Media survey of more than 1,200 kids aged 9 to 17, reported in June 2026, found over a third of AI users had talked about feelings or personal problems with a chatbot, and 57% had asked for health or body advice. What that does to kids over time is not yet known; there are no long-term studies. Chatbots are not licensed clinicians or crisis services.

Verdict: the concern is reasonable; the evidence on effects is thin. The last example, knowing whom to contact, is the one non-negotiable.

For a 7-year-old: “When you feel sad or scared, you come to Mom, Dad or your teacher.”

For a 14-year-old: Write the crisis contacts into the agreement itself: a parent, another trusted adult, and in the US the 988 Suicide & Crisis Lifeline (call or text 988) or Crisis Text Line (text HOME to 741741). Outside the US, find a line at findahelpline.com.

Section 4: Sleep

The template asks: “What guidelines should there be for safeguarding sleep?” It suggests screen-free sleep with devices out of bedrooms, or, if devices are allowed, how to protect sleep anyway.

What the research says. This has the strongest evidence of any section. A 2016 meta-analysis in JAMA Pediatrics by Ben Carter and colleagues pooled 20 studies of 125,198 children aged 6 to 19. Bedtime media device use was linked to more than double the odds of inadequate sleep (OR 2.17) and higher odds of poor sleep quality and daytime sleepiness. The striking part: children who merely had access to a device at night, without using it, also slept worse (OR 1.79 for inadequate sleep). The studies were cross-sectional, so they show association, not cause, but the pattern was consistent.

Verdict: strongly supported, and not AI-specific. A chatbot that is always awake is one more reason to keep the phone outside the bedroom.

For a 7-year-old: “Tablets sleep in the kitchen.”

For a 14-year-old: “Phone charges outside your room on school nights.” Negotiate weekends. Model it yourself.

Section 5: Fakes

The template asks: “What guidelines should there be regarding AI-generated content?” Examples: creating or sharing fake photos and videos, altering someone’s appearance or cloning their voice without permission, using AI for insults, rumors or impersonation, and sharing suspicious images without verifying.

What the research says. The legal stakes are now clear in the US. The federal Take It Down Act requires covered platforms to remove non-consensual intimate images, and known identical copies, within 48 hours of a valid request; the Federal Trade Commission began enforcing that on May 19, 2026. Creating or sharing a fake intimate image of a classmate is not a prank. It can be a crime.

Verdict: well grounded. This section protects your kid as a potential target and as a potential bystander or creator, and both roles need talking about.

For a 7-year-old: “We don’t change pictures of people to make fun of them. Ask before taking anyone’s photo.”

For a 14-year-old: “Never make or forward a fake image of a real person without their permission. If you see one of someone you know, tell me, and don’t share it.” Explain that forwarding an explicit image of a minor can itself be illegal.

Section 6: Privacy

The template asks: “What guidelines should there be regarding sharing personal info with AI?” Examples: full names, addresses, phone numbers, school, passwords; personal photos, private conversations, health and emotional information; private details about friends and family.

What the research says. Common Sense Media’s 2025 report found about a quarter of teens had shared personal information with AI companion platforms. What a chatbot is told is usually stored on company servers and, if the product has memory, can resurface in later chats.

Verdict: well grounded and very practical. It is also the easiest section to turn into a checklist.

For a 7-year-old: “Never tell a computer your full name, address or school.”

For a 14-year-old: “No addresses, passwords or photos of friends. Check whether an app remembers chats and turn it off if you can.” Walk through one app’s settings together.

The Safety-First Commitment: The Part Parents Sign

The template ends with a pledge for parents: to be someone their child can turn to “when things go awry online,” and to take “a team approach” rather than “reflexively taking away tech or some other punishment.” It notes taking tech away may sometimes be necessary, “but that should not be the first response.”

This has a solid research backbone. In a widely cited 2000 study in Developmental Psychology, Margaret Kerr and Håkan Stattin followed 1,186 14-year-olds in Sweden. Parents’ knowledge of their kids’ lives was linked to better adjustment, but most of that knowledge came from children choosing to tell, not from parents tracking and surveilling. Kids who felt controlled did worse. Translated to AI: you will mostly know what happens online because your kid tells you, and they tell you when telling feels safe.

The Agreement at a Glance: Evidence and Age Versions

Section	Strength of evidence	Key source	Rule for a 7-year-old	Rule for a 14-year-old
Social life	Moderate (use well documented; effects unclear)	Common Sense 2025; Liu & Ishfaq 2026 preprint	Chatbots aren’t friends; tell a grown-up about secrets	Notice if AI starts replacing friends
Academic life	Strong	Bastani et al., PNAS 2025	Ask first before AI homework	Try first; AI explains, never does
Mental wellness	Thin on effects; clear on crisis care	Common Sense 2026 survey	Feelings go to a grown-up	Crisis contacts written in the agreement
Sleep	Strong (consistent, cross-sectional)	Carter et al., JAMA Pediatrics 2016	Tablets sleep in the kitchen	Phone charges outside the room
Fakes	Strong legal basis	FTC, Take It Down Act enforcement 2026	Don’t change people’s pictures	Never make or forward fakes of real people
Privacy	Moderate to strong	Common Sense 2025	Never share name, address, school	No personal details; turn off chat memory
Safety-first pledge	Strong	Kerr & Stattin, Developmental Psychology 2000	”You can always tell me”	Team response before taking tech

How to Fill It In Without a Fight

Do one section per sitting

Six sections in one night is a lecture. Start with the most urgent (often sleep for young kids, social life or school for teens) and do one a week.

Let the teen write the first draft

Hand a 14-year-old the pen. Rules kids draft are rules they can explain. Ask “why?” instead of rewriting.

Write the why next to the rule

The page’s own advice is to focus on “WHYs.” A rule with its reason attached (“phone outside the room because sleep research is clear”) survives arguments better than a bare rule.

Set the review date before you sign

Fill in the template’s “Date to Revisit and Review” line. AI products change monthly.

What not to do

Don’t use the agreement as a contract to enforce with punishments. The template itself says perfect rule-following isn’t the goal. And don’t sign the safety-first pledge if you’re not ready to keep it; breaking it once teaches your kid not to tell you next time.

What to Watch For Over the Next 3 Months

• Week 4: Is the agreement on the fridge or in a drawer? Have you finished at least two sections?
• Month 2 red flags: rules followed only when you’re watching, a sudden drop in what your child tells you about online life, or a broken pledge on your side (taking the phone away as a first reaction).
• Month 3 self-check: Hold the review. Which rule is ignored, and is it the wrong rule? Which section needs updating because a new app or feature appeared? For questions to restart the conversation, see our discussion guide for after Screenagers: Generation AI, and for background on the film, our overview of what Screenagers: Generation AI covers.

Frequently Asked Questions

What is a family AI agreement?

A family AI agreement is a short written plan for how a household uses AI chatbots and tools, with reasons attached. The free Screenagers version covers social life, schoolwork, mental wellness, sleep, AI fakes and privacy, and ends with a parent pledge to respond to problems as a team.

At what age should we make a family AI agreement?

As soon as your child uses any talking device or app, but the content should change with age. For a 7-year-old, turn each section into one simple house rule. For a 13- or 14-year-old, let them draft the rules. Revisit it every few months, as the template recommends.

Does the research support banning AI at home?

Not as a blanket rule. The evidence supports specific limits: devices out of bedrooms at night, AI as explainer rather than answer-machine for homework, and no sharing of personal information. A 2025 math experiment found guardrailed AI tutoring avoided the learning losses that unrestricted use caused.

What if my child breaks the agreement?

Treat it as information, not betrayal. The template’s own goal is conversation about the why, not perfect compliance. Ask what made the rule hard to follow, and adjust if the rule is wrong. Research on teen disclosure suggests a calm response keeps kids telling you things.

---

About the author

Ricky Flores is the founder of HiWave Makers and an electrical engineer with 15+ years of experience building consumer technology at Apple, Samsung, and Texas Instruments. He writes about how kids learn to build, think, and create in a tech-saturated world. Read more at hiwavemakers.com.

---

Sources

1. Screenagers. (2026, October 27). “Our Family AI Agreement for Kids and Teens” (with downloadable PDF, “Family AI Agreement For Kids & Teens”). https://www.screenagersmovie.com/blog/family-ai-agreement
2. American Academy of Pediatrics. (2026, January 20). “A Child-Friendly Digital World: AAP Releases New Media Recommendations.” HealthyChildren.org. https://www.healthychildren.org/English/news/Pages/creating-a-child-friendly-digital-world-AAP-releases-new-media-recommendations.aspx
3. Common Sense Media. (2025, July 16). “Talk, Trust, and Trade-Offs: How and Why Teens Use AI Companions.” https://www.commonsensemedia.org/research/talk-trust-and-trade-offs-how-and-why-teens-use-ai-companions
4. Liu, J., & Ishfaq, N. (2026). “‘I Talked an AI Chatbot, So What’s Next?’ How U.S. Young Adults Imagine Responsible AI for Emotion Coping.” arXiv (preprint, not peer-reviewed). https://arxiv.org/abs/2609.25311
5. Bastani, H., Bastani, O., Sungu, A., Ge, H., Kabakcı, Ö., & Mariman, R. (2025). “Generative AI without guardrails can harm learning: Evidence from high school mathematics.” Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2422633122
6. Smith, C. (2026, June 8). “Nearly 9 in 10 kids use AI; report flags concerns over educational, emotional dependencies” (Common Sense Media survey). The National News Desk / KOKH. https://okcfox.com/news/nation-world/nearly-9-in-10-kids-use-ai-report-flags-concerns-over-educational-emotional-dependency-common-sense-media-youth-ai-safety-institute-common-sense-media-census-ai-use-by-tweens-and-teens-2026
7. Carter, B., Rees, P., Hale, L., Bhattacharjee, D., & Paradkar, M. S. (2016). “Association Between Portable Screen-Based Media Device Access or Use and Sleep Outcomes: A Systematic Review and Meta-analysis.” JAMA Pediatrics. https://doi.org/10.1001/jamapediatrics.2016.2341
8. Federal Trade Commission. (2026, May 19). “FTC begins enforcing TAKE IT DOWN Act.” https://www.ftc.gov/news-events/news/press-releases/2026/05/ftc-begins-enforcing-take-it-down-act
9. Kerr, M., & Stattin, H. (2000). “What parents know, how they know it, and several forms of adolescent adjustment: Further support for a reinterpretation of monitoring.” Developmental Psychology, 36(3). https://doi.org/10.1037/0012-1649.36.3.366