Does ChatGPT have parental controls, and what can I actually see?

Yes. Since September 2025 parents can link accounts with a teen, set quiet hours, turn off memory, voice and image tools, opt out of training, and get alerts about possible self-harm. You cannot read your teen's chats, and independent testers found the alerts and age detection unreliable, so the controls are a backstop, not a substitute for talking.

Can my child under 13 use Google Gemini?

In many countries, yes, if a parent allows it through a supervised Google account in Family Link, where you can switch it on or off. Kids get extra filters, but Google says they are not perfect, and the option is not offered to supervised kids in the UK or EU.

Can teens still chat with Character.AI or Meta's AI characters?

Character.AI ended open-ended chat for under-18s in late 2025, leaving creative tools only. Meta paused teen access to its AI characters in January 2026 while it builds a version with parent controls; Meta AI itself stays available with teen protections.

Which parental controls should I set up on my child's phone or tablet?

Start with the free tools built into the device: Apple Screen Time and Family Sharing, Google Family Link, Microsoft Family Safety, or the Amazon Parent Dashboard. Surveys suggest families who use them mostly find they work, but experts stress pairing controls with clear family rules and regular conversations.

What does COPPA protect, and what changed in 2025?

COPPA requires apps and sites aimed at kids under 13, or that know a child is using them, to get a parent's consent before collecting personal data, and lets you review or delete it. The 2025 update, with full compliance due April 2026, added separate consent for targeted advertising and limits on how long data is kept.

Is the Kids Online Safety Act (KOSA) law yet?

Not as of late October 2026. The Senate Commerce Committee sent an amended KOSA to the full Senate in August 2026, and the House passed its own KIDS Act in June 2026 without KOSA's duty of care, so the two chambers still have to agree.

Are there laws about AI companion chatbots and kids?

A few US states have acted. California's SB 243 (2025) requires AI disclosure, break reminders for minors, limits on sexual content and self-harm protocols, and New York, Maine and Utah have disclosure rules. The FTC is also studying how companion bots affect kids. Protections differ by state and are still new.

How do I know if a learning app is safe for my child's privacy?

Look it up in the Common Sense Privacy Program, which rates kids' and school apps on their privacy policies, and check whether it asks for parental consent if your child is under 13. Ratings judge what a company promises, not what it actually does, so keep settings tight and share little.

What rights do I have over my child's school records and the apps their school uses?

Under FERPA you can see and ask to correct your child's education records, and schools generally need consent before sharing them, with exceptions for vendors acting as the school. It is reasonable to ask the school which apps and AI tools collect student data and how long they keep it.

What does my state tell schools about using AI?

About 35 states plus Puerto Rico had issued official K-12 AI guidance by August 2026, usually covering privacy, academic honesty and teacher training; a few states leave it to districts. Some advocacy groups also offer model policies parents can bring to a school board.

What is Australia's under-16 social media rule and does it cover AI apps?

Since 10 December 2025, major platforms such as TikTok, Instagram, Snapchat and YouTube must take reasonable steps to stop Australians under 16 from having accounts, and parents cannot opt a child in. Messaging, gaming, YouTube Kids and education tools are not covered, and AI chatbots are not on the list as of the latest FAQ.

Do age checks actually work?

Partly. Australia's government trial said age assurance is technically workable, but face-scanning estimates were often off by more than a year for teens, and some vendors collected too much data. Thorn found nearly half of US minors had used an app before its minimum age. Expect age checks to reduce, not stop, underage use.

What do the UK and EU rules mean for kids' apps?

In the UK, the Children's Code requires high-privacy defaults for under-18s, and the Online Safety Act children's codes (July 2025) require age checks and safer feeds. In the EU, the DSA minors guidelines push similar defaults, and the AI Act bans AI that exploits children's age and treats AI that grades or places students as high-risk.

A nude image of my child, real or AI-made, is being shared or used as a threat. What can I do?

Use NCMEC's free Take It Down tool, which fingerprints the image on your device so participating platforms can remove copies without you uploading it, and report to NCMEC's CyberTipline or local police. Reassure your child they are not in trouble; shame often keeps kids from asking for help.

Which online safety organizations can I trust, and who funds them?

Government regulators (eSafety, ICO, Ofcom) and child charities (NSPCC, 5Rights, Fairplay, Thorn) give practical advice with fewer commercial ties. Internet Matters, ConnectSafely and FOSI publish useful how-to guides but are partly funded by or count as members the tech companies they write about, so read them as a start.

The Big Index

Safety, privacy and the rules

The short version: parental controls on AI tools, privacy ratings, and the laws that protect kids online.

50 entries
49 free
10 must-reads ★

Questions parents ask

Does ChatGPT have parental controls, and what can I actually see?
Yes. Since September 2025 parents can link accounts with a teen, set quiet hours, turn off memory, voice and image tools, opt out of training, and get alerts about possible self-harm. You cannot read your teen's chats, and independent testers found the alerts and age detection unreliable, so the controls are a backstop, not a substitute for talking.
- Introducing parental controls (OpenAI)
- Parents' Ultimate Guide to ChatGPT Parental Controls (Common Sense Media)
- ChatGPT claims to have a safety net for teens. We watched it fail. (Common Sense Media)
More in Safety, privacy and the rules
Can my child under 13 use Google Gemini?
In many countries, yes, if a parent allows it through a supervised Google account in Family Link, where you can switch it on or off. Kids get extra filters, but Google says they are not perfect, and the option is not offered to supervised kids in the UK or EU.
- Guiding your child's Gemini Apps experience (Google)
- Family Link (Google)
More in Safety, privacy and the rules
Can teens still chat with Character.AI or Meta's AI characters?
Character.AI ended open-ended chat for under-18s in late 2025, leaving creative tools only. Meta paused teen access to its AI characters in January 2026 while it builds a version with parent controls; Meta AI itself stays available with teen protections.
- Taking Bold Steps to Keep Teen Users Safe on Character.AI (Character.AI)
- Empowering Parents, Protecting Teens: Meta's Approach to AI Safety (Meta)
More in Safety, privacy and the rules
Which parental controls should I set up on my child's phone or tablet?
Start with the free tools built into the device: Apple Screen Time and Family Sharing, Google Family Link, Microsoft Family Safety, or the Amazon Parent Dashboard. Surveys suggest families who use them mostly find they work, but experts stress pairing controls with clear family rules and regular conversations.
- Set up parental controls to manage your child's iPhone or iPad (Apple)
- Family Link (Google)
- Microsoft Family Safety (Microsoft)
- Amazon Parent Dashboard (Amazon)
- New FOSI Research: Parents Are Underestimating Their Kids' Digital Lives - Especially With AI (Family Online Safety Institute and Ipsos)
More in Safety, privacy and the rules
What does COPPA protect, and what changed in 2025?
COPPA requires apps and sites aimed at kids under 13, or that know a child is using them, to get a parent's consent before collecting personal data, and lets you review or delete it. The 2025 update, with full compliance due April 2026, added separate consent for targeted advertising and limits on how long data is kept.
- Protecting Your Child's Privacy Online (Federal Trade Commission)
- Complying with COPPA: Frequently Asked Questions (Federal Trade Commission)
Is the Kids Online Safety Act (KOSA) law yet?
Not as of late October 2026. The Senate Commerce Committee sent an amended KOSA to the full Senate in August 2026, and the House passed its own KIDS Act in June 2026 without KOSA's duty of care, so the two chambers still have to agree.
- Kids Online Safety Act (CRS In Focus IF12730) (Congressional Research Service)
- Kids Online Safety Act (Wikipedia)
Are there laws about AI companion chatbots and kids?
A few US states have acted. California's SB 243 (2025) requires AI disclosure, break reminders for minors, limits on sexual content and self-harm protocols, and New York, Maine and Utah have disclosure rules. The FTC is also studying how companion bots affect kids. Protections differ by state and are still new.
- Understanding the New Wave of Chatbot Legislation: California SB 243 and Beyond (Justine Gluck, Future of Privacy Forum)
- SB-243 Companion chatbots (California Legislature)
- FTC launches inquiry into AI chatbots acting as companions (Federal Trade Commission)
- Artificial Intelligence 2025 Legislation (National Conference of State Legislatures)
How do I know if a learning app is safe for my child's privacy?
Look it up in the Common Sense Privacy Program, which rates kids' and school apps on their privacy policies, and check whether it asks for parental consent if your child is under 13. Ratings judge what a company promises, not what it actually does, so keep settings tight and share little.
- Common Sense Privacy Program (Common Sense Media)
- 2023 State of Kids' Privacy (Common Sense Media)
- Protecting Your Child's Privacy Online (Federal Trade Commission)
What rights do I have over my child's school records and the apps their school uses?
Under FERPA you can see and ask to correct your child's education records, and schools generally need consent before sharing them, with exceptions for vendors acting as the school. It is reasonable to ask the school which apps and AI tools collect student data and how long they keep it.
- Understanding Your Child's School Records: A Parent's Primer on FERPA (US Department of Education)
- Ask Your Kid's School These Essential Student Privacy and Safety Questions (Common Sense Media)
- State AI Guidance for K12 Schools (AI for Education)
What does my state tell schools about using AI?
About 35 states plus Puerto Rico had issued official K-12 AI guidance by August 2026, usually covering privacy, academic honesty and teacher training; a few states leave it to districts. Some advocacy groups also offer model policies parents can bring to a school board.
- State AI Guidance for K12 Schools (AI for Education)
- Student-First AI Policy Offers Model for Districts Nationwide (Fairplay)
What is Australia's under-16 social media rule and does it cover AI apps?
Since 10 December 2025, major platforms such as TikTok, Instagram, Snapchat and YouTube must take reasonable steps to stop Australians under 16 from having accounts, and parents cannot opt a child in. Messaging, gaming, YouTube Kids and education tools are not covered, and AI chatbots are not on the list as of the latest FAQ.
- Social media age restrictions and your family (eSafety Commissioner (Australia))
- Social media 'ban' or delay FAQs (eSafety Commissioner (Australia))
- Online Safety Amendment (Social Media Minimum Age) Act 2024 (Wikipedia)
Do age checks actually work?
Partly. Australia's government trial said age assurance is technically workable, but face-scanning estimates were often off by more than a year for teens, and some vendors collected too much data. Thorn found nearly half of US minors had used an app before its minimum age. Expect age checks to reduce, not stop, underage use.
- Trial finds age assurance can be done, as under-16s social media ban deadline looms (ABC News (Australia))
- Youth Perspectives on Online Safety, 2025 (Thorn)
- Building towards age prediction (OpenAI)
- Commission presents guidelines and age verification app prototype for a safer online space for children (European Commission)
What do the UK and EU rules mean for kids' apps?
In the UK, the Children's Code requires high-privacy defaults for under-18s, and the Online Safety Act children's codes (July 2025) require age checks and safer feeds. In the EU, the DSA minors guidelines push similar defaults, and the AI Act bans AI that exploits children's age and treats AI that grades or places students as high-risk.
- The Children's code and children's online privacy (Information Commissioner's Office (UK))
- New rules for a safer generation of children online (Ofcom)
- Commission publishes guidelines on the protection of minors (European Commission)
- Article 5: Prohibited AI Practices (Future of Life Institute (EU AI Act explorer))
- Annex III: High-Risk AI Systems Referred to in Article 6(2) (Future of Life Institute (EU AI Act explorer))
A nude image of my child, real or AI-made, is being shared or used as a threat. What can I do?
Use NCMEC's free Take It Down tool, which fingerprints the image on your device so participating platforms can remove copies without you uploading it, and report to NCMEC's CyberTipline or local police. Reassure your child they are not in trouble; shame often keeps kids from asking for help.
- Take It Down (National Center for Missing & Exploited Children)
- NetSmartz (National Center for Missing & Exploited Children)
- Youth Perspectives on Online Safety, 2025 (Thorn)
Which online safety organizations can I trust, and who funds them?
Government regulators (eSafety, ICO, Ofcom) and child charities (NSPCC, 5Rights, Fairplay, Thorn) give practical advice with fewer commercial ties. Internet Matters, ConnectSafely and FOSI publish useful how-to guides but are partly funded by or count as members the tech companies they write about, so read them as a start.
- 5Rights Foundation (5Rights Foundation)
- Internet Matters (Internet Matters)
- ConnectSafely (ConnectSafely)
- Family Online Safety Institute (FOSI) (Family Online Safety Institute)
- Fairplay (Fairplay (formerly Campaign for a Commercial-Free Childhood))

Must-reads

AppFree
Family Link
Google
Google's free parental control app for Android phones, Chromebooks and Google accounts: screen time limits, app approvals, content filters for YouTube, Chrome and Search, and location sharing. Also where you manage a child's access to Gemini.
families.google
ToolFree
Common Sense Privacy Program
Common Sense Media
Free privacy ratings of thousands of kids' apps and school tools, based on reading each company's privacy policy. Check an app before your child signs up, or ask your school whether its tools passed. Ratings judge policies, not actual data practices.
privacy.commonsense.org
GuideFree
Protecting Your Child's Privacy Online
Federal Trade Commission
The FTC's parent-facing page on COPPA, the US law that makes apps and sites get a parent's consent before collecting data from kids under 13. Explains what consent requests look like and what you can ask a company to delete.
consumer.ftc.gov
ArticleFree
Understanding the New Wave of Chatbot Legislation: California SB 243 and Beyond
Justine Gluck, Future of Privacy Forum, 2025
Clear rundown of the first US laws on AI companion chatbots: California SB 243, New York's companion law, plus Maine and Utah disclosure rules and bills on therapy bots. Good map of what protections now exist where you live. FPF has industry supporters.
fpf.org
GuideFree
Switched on to privacy
Information Commissioner's Office (UK), 2026
A 2026 UK campaign for parents of kids aged 4 to 11 with short videos, real parent stories and a three-step routine: chat about privacy, choose what to share, and check settings each time a new app or device arrives.
ico.org.uk
GuideFree
Social media age restrictions and your family
eSafety Commissioner (Australia), 2025
Australia's official parent page for the under-16 social media minimum age that began 10 December 2025. Explains which apps are covered, what happens to existing accounts, and how to keep talking with kids who may try to get around it.
esafety.gov.au
GuideFree
Parents' Ultimate Guide to ChatGPT Parental Controls
Common Sense Media, 2026
Step-by-step setup with a suggested setting for each switch (sensitive content on, model training off, image tools off, alerts on). Honest about the gaps: teens can make an unlinked account, and alerts are not reliable in a crisis.
commonsensemedia.org
GuideFree
Guiding your child's Gemini Apps experience
Google, 2025
How parents turn Gemini on or off for a child under 13 in Family Link, and what extra filters kids get. Google admits filters are imperfect and suggests using it together. Not offered to supervised kids in the UK or EU.
support.google.com
GuideFree
Set up parental controls to manage your child's iPhone or iPad
Apple
Apple's guide to Screen Time and Family Sharing: Ask to Buy, website limits, contact limits, downtime schedules, age ratings, and Communication Safety, which blurs nude images. Many protections switch on by default for child accounts.
support.apple.com
GuideFree
Understanding Your Child's School Records: A Parent's Primer on FERPA
US Department of Education, 2021
Official plain-English guide to FERPA, the federal law that lets parents see and correct their child's school records and limits who schools can share them with. Rights pass to the student at 18. In English and Spanish.
studentprivacy.ed.gov

All 50

Device and account controls

★Family LinkGoogle · App · FreeGoogle's free parental control app for Android phones, Chromebooks and Google accounts: screen time limits, app approvals, content filters for YouTube, Chrome and Search, and location sharing. Also where you manage a child's access to Gemini.
★Set up parental controls to manage your child's iPhone or iPadApple · Guide · FreeApple's guide to Screen Time and Family Sharing: Ask to Buy, website limits, contact limits, downtime schedules, age ratings, and Communication Safety, which blurs nude images. Many protections switch on by default for child accounts.
Microsoft Family SafetyMicrosoft · App · Free + paidFree tool for Windows, Xbox and Android: screen time limits by app and game, activity reports, and web filters. Kids can ask for more time from their device. Web filtering works only in the Edge browser, so other browsers get blocked.
Parents' Ultimate Guide to Parental ControlsCommon Sense Media · 2021 · Guide · FreeA broad overview of every kind of control, from built-in phone settings to router filters and monitoring apps, with honest notes on what each can and cannot do. Some product details are dated (first posted 2021), but the framework holds.
Parental controlsNSPCC · Guide · FreeThe UK children's charity's practical page on setting controls on home broadband, phones, consoles and apps, with the reminder that controls work best alongside regular talks about what kids see online.

Privacy and school records

★Common Sense Privacy ProgramCommon Sense Media · Tool · FreeFree privacy ratings of thousands of kids' apps and school tools, based on reading each company's privacy policy. Check an app before your child signs up, or ask your school whether its tools passed. Ratings judge policies, not actual data practices.
2023 State of Kids' PrivacyCommon Sense Media · 2023 · Report · FreeCommon Sense's review of popular apps found many do not clearly say how they make money from user data, and argued privacy laws need real enforcement. Background for why reading a privacy rating matters.
Ask Your Kid's School These Essential Student Privacy and Safety QuestionsCommon Sense Media · 2018 · Guide · FreeA short list of questions to bring to a school meeting: which apps collect student data, who can see it, and how long it is kept. Written before the AI boom, but the same questions apply to AI tutors and chatbots used in class.
★Understanding Your Child's School Records: A Parent's Primer on FERPAUS Department of Education · 2021 · Guide · FreeOfficial plain-English guide to FERPA, the federal law that lets parents see and correct their child's school records and limits who schools can share them with. Rights pass to the student at 18. In English and Spanish.
Student-First AI Policy Offers Model for Districts NationwideFairplay · 2026 · Guide · FreeAn October 2026 model school policy from an advocacy group skeptical of AI in classrooms: keep AI products out unless independent evidence shows better learning, strong privacy and no addictive design. Parents can bring it to a school board.

US law

★Protecting Your Child's Privacy OnlineFederal Trade Commission · Guide · FreeThe FTC's parent-facing page on COPPA, the US law that makes apps and sites get a parent's consent before collecting data from kids under 13. Explains what consent requests look like and what you can ask a company to delete.
Complying with COPPA: Frequently Asked QuestionsFederal Trade Commission · Guide · FreeThe FTC's long Q&A for companies on the children's privacy rule. Written for businesses, but shows what an app must do: notify parents, get consent, let you delete data. The 2025 update (full compliance by April 2026) added separate consent for targeted ads and data-retention limits.
Kids Online Safety ActWikipedia · 2026 · Article · Free · TeensReadable history of KOSA, the federal bill that would give platforms a duty to protect minors and give parents more settings. It passed the Senate 91-3 in 2024, then stalled in the House. As of late 2026 it is still not law.
Kids Online Safety Act (CRS In Focus IF12730)Congressional Research Service · 2026 · Report · FreeNeutral two-page summary from Congress's research office, updated August 2026. Lists what KOSA would require (parent tools, limits on compulsive design, audits) and compares the Senate bill, sent to the floor in August 2026, with the House's KIDS Act, passed June 2026 without a duty of care.

AI chatbot rules

FTC launches inquiry into AI chatbots acting as companionsFederal Trade Commission · 2025 · Report · Free · TeensSeptember 2025 press release: the FTC ordered several large AI and social media companies to explain how their companion chatbots affect kids and teens, how they test for harm, and what they tell parents. A study, not a rule, but a sign of where US oversight is heading.
★Understanding the New Wave of Chatbot Legislation: California SB 243 and BeyondJustine Gluck, Future of Privacy Forum · 2025 · Article · FreeClear rundown of the first US laws on AI companion chatbots: California SB 243, New York's companion law, plus Maine and Utah disclosure rules and bills on therapy bots. Good map of what protections now exist where you live. FPF has industry supporters.
SB-243 Companion chatbotsCalifornia Legislature · 2025 · Law · Free · TeensCalifornia's companion chatbot law, signed October 2025. Bots must say they are AI, remind known minors to take a break every three hours, block sexual content for minors, and follow a self-harm protocol. A stricter bill, AB 1064, was vetoed.
Artificial Intelligence 2025 LegislationNational Conference of State Legislatures · 2025 · Data · FreeA searchable state-by-state table of AI bills from the 2025 sessions; every state introduced some and 38 enacted around 100 measures. Useful to look up whether your state passed anything on chatbots, schools or minors.

UK and EU law

The Children's code and children's online privacyInformation Commissioner's Office (UK) · Guide · FreePlain-English page for families on the UK Age Appropriate Design Code. Explains what apps used by under-18s must now do, such as high-privacy defaults, location off, and no nudges to share more, and how to complain if they do not.
Age appropriate design: a code of practice for online servicesInformation Commissioner's Office (UK) · 2020 · Law · FreeThe full Children's Code: 15 standards in force since 2021 for any online service likely to be used by kids, from games to learning apps. Its 'best interests of the child' approach has been copied by other countries and US states.
★Switched on to privacyInformation Commissioner's Office (UK) · 2026 · Guide · Free · Ages 5–8A 2026 UK campaign for parents of kids aged 4 to 11 with short videos, real parent stories and a three-step routine: chat about privacy, choose what to share, and check settings each time a new app or device arrives.
New rules for a safer generation of children onlineOfcom · 2025 · Guide · Free · TeensOfcom's summary of the UK Online Safety Act children's codes, in force since July 2025. Platforms must check ages before showing porn or self-harm content, tame kids' feeds, and give children easy reporting. Explains what families should now expect to see.
Online Safety Act: explainerUK Government (DSIT) · 2025 · Guide · FreeThe government's overview of the Online Safety Act: which content counts as harmful to children, the July 2025 deadlines, and fines up to 10% of global revenue. Says little about AI chatbots specifically.
Commission publishes guidelines on the protection of minorsEuropean Commission · 2025 · Guide · FreeThe EU's July 2025 guidelines under the Digital Services Act on what platforms open to minors should do: private accounts by default, safer recommendations, limits on addictive features and loot boxes, and reliable age checks. Non-binding but used to judge compliance.
The Digital Services Act (DSA) explained: what online platforms should do to keep kids and teens safe onlineEuropean Commission · 2025 · Guide · Free · TeensA short family-friendly booklet, in every EU language, that turns the DSA minors guidelines into plain words. Handy for parents and older kids in Europe who want to know what to expect from apps and how to report problems.
Protecting and empowering young people onlineEuropean Commission · 2026 · Guide · FreeThe EU's hub page on kids' online safety. Collects the DSA guidelines, the age-verification app, Better Internet for Kids resources, and news of the KIDS Act proposed in September 2026, which would add a minimum age for social media.
Article 5: Prohibited AI PracticesFuture of Life Institute (EU AI Act explorer) · 2024 · Law · FreeReadable copy of the EU AI Act's banned uses, including AI that exploits a child's age to distort behavior and emotion recognition in schools. These bans have applied since February 2025. The site is run by a nonprofit, not the EU.
Annex III: High-Risk AI Systems Referred to in Article 6(2)Future of Life Institute (EU AI Act explorer) · 2024 · Law · FreeThe list of 'high-risk' AI uses under the EU AI Act. Section 3 covers education: AI that decides admissions, grades work, places students by level, or watches for cheating. Those tools face testing and oversight rules, though the start date has been pushed back.
EU AI Act: How Well Does it Protect Children and Young People?Nomisha Kurian, Leverhulme Centre for the Future of Intelligence · 2024 · Blog · FreeA Cambridge researcher's short take on what the AI Act gets right for kids (explicit child-rights language) and where it leaves gaps for children. Written in April 2024, before the final text, so a few details have moved.

Australia and age checks

★Social media age restrictions and your familyeSafety Commissioner (Australia) · 2025 · Guide · Free · TeensAustralia's official parent page for the under-16 social media minimum age that began 10 December 2025. Explains which apps are covered, what happens to existing accounts, and how to keep talking with kids who may try to get around it.
Social media 'ban' or delay FAQseSafety Commissioner (Australia) · 2025 · Guide · Free · TeensDetailed answers on Australia's under-16 rule: the list of restricted platforms (TikTok, Instagram, YouTube, Snapchat and others), what stays open (YouTube Kids, Roblox, messaging, school tools), and why parents cannot opt their child in.
Online Safety Amendment (Social Media Minimum Age) Act 2024Wikipedia · 2026 · Article · FreeBackground on Australia's under-16 law: how it passed, the age-assurance trial, criticism from privacy and youth groups, and how platforms responded after December 2025. A balanced starting point before reading the official pages.
Trial finds age assurance can be done, as under-16s social media ban deadline loomsABC News (Australia) · 2025 · Article · FreeReport on the government's age-checking technology trial. Testers said age checks are workable, but face-scanning was off by more than 18 months for some teens and some vendors over-collected data. Good context on how accurate age checks really are.
Building towards age predictionOpenAI · 2025 · Article · Free · TeensOpenAI's September 2025 plan to estimate whether a ChatGPT user is under 18 from how they use it, default unsure users to the teen experience, and ask for ID in some places. Written by the company, so read alongside independent tests.
Commission presents guidelines and age verification app prototype for a safer online space for childrenEuropean Commission · 2025 · Article · FreeJuly 2025 announcement of the EU's prototype age-verification app, which lets someone prove they are over 18 without sharing other personal details. A test of 'privacy-preserving' age checks being piloted in several EU countries.

Safety organizations

Children & AI Design Code5Rights Foundation · 2025 · Guide · FreeA March 2025 checklist for anyone building AI that children might use, written with child-rights and AI experts. Parents can borrow its questions (Is it safe by default? Who is accountable?) when sizing up an AI app.
5Rights Foundation5Rights Foundation · Organization · FreeUK-based charity founded by Beeban Kidron that pushed for the UK Children's Code and now works on AI and children worldwide. Publishes research on harmful design in apps, games and AI, and policy toolkits.
Internet MattersInternet Matters · Organization · FreeUK nonprofit with step-by-step parental control guides for hundreds of devices and apps, plus research on kids and AI chatbots. Founded and funded by internet providers and tech firms including Google and Meta, which is worth knowing when reading its advice.
Talking to children about AINSPCC · Guide · FreeThe UK child protection charity's advice on starting conversations about AI chatbots and image tools, what risks to watch for, and where to get help. The NSPCC also runs a helpline and the Report Remove tool with Childline.
ConnectSafelyConnectSafely · Organization · FreeUS nonprofit with short 'Parent's Guide' PDFs on apps, games and generative AI, plus a podcast. Practical and calm in tone. Several guides are produced with funding from the companies they cover, including Meta, so treat them as a start.
Family Online Safety Institute (FOSI)Family Online Safety Institute · Organization · FreeUS nonprofit with free parent resources: 7 Steps to Good Digital Parenting, AI conversation starters, a family online safety agreement, and guides to Roblox and YouTube. Its members include tech companies such as OpenAI, TikTok, Roblox and Snap.
New FOSI Research: Parents Are Underestimating Their Kids' Digital Lives - Especially With AIFamily Online Safety Institute and Ipsos · 2026 · Study · Free · Teens2026 survey of 4,000+ US and Australian families with kids 10-17. Kids reported more AI use than parents knew (38% vs 27% in the past week). Families used rules (68%) more than parental controls (49%), but most who used controls said they worked. Industry-funded group.
Youth Perspectives on Online Safety, 2025Thorn · 2026 · Study · Free · TeensThorn's survey of about 1,000 US minors: 47% used an app before its minimum age, 67% used AI chatbots or companions, and some turned to AI for advice after a bad online sexual experience. Now tracks deepfake nudes and sextortion. From a nonprofit that builds child-safety tech.

Parental controls in AI tools

Introducing parental controlsOpenAI · 2025 · Article · Free · TeensOpenAI's September 2025 launch of ChatGPT parental controls. Parents link accounts with a teen (13+) and can set quiet hours, turn off memory, voice and image tools, opt out of training, and get alerts if a chat suggests self-harm risk. Parents cannot read chats.
★Parents' Ultimate Guide to ChatGPT Parental ControlsCommon Sense Media · 2026 · Guide · Free · TeensStep-by-step setup with a suggested setting for each switch (sensitive content on, model training off, image tools off, alerts on). Honest about the gaps: teens can make an unlinked account, and alerts are not reliable in a crisis.
ChatGPT claims to have a safety net for teens. We watched it fail.Common Sense Media · 2026 · Report · Free · TeensCommon Sense tested ChatGPT's teen mode with thousands of prompts in 2026 and found teens often went unrecognized, parent alerts did not fire in crisis chats, and hotline referrals dropped. It rates ChatGPT an unacceptable risk for kids. An advocacy test, not peer reviewed.
Taking Bold Steps to Keep Teen Users Safe on Character.AICharacter.AI · 2025 · Article · Free · TeensCharacter.AI's October 2025 announcement that under-18 users would lose open-ended chat with characters by late November 2025, keeping only creative tools, plus new age checks. A company post that followed news reports, lawsuits and regulator questions.
Empowering Parents, Protecting Teens: Meta's Approach to AI SafetyMeta · 2025 · Article · Free · TeensMeta's plan for parents to turn off teens' one-on-one chats with AI characters, block specific characters, and see topics discussed. A January 2026 update paused teen access to AI characters while a new version is built; Meta AI itself stays on. Company post.
★Guiding your child's Gemini Apps experienceGoogle · 2025 · Guide · Free · Ages 8–12How parents turn Gemini on or off for a child under 13 in Family Link, and what extra filters kids get. Google admits filters are imperfect and suggests using it together. Not offered to supervised kids in the UK or EU.

Getting help

NetSmartzNational Center for Missing & Exploited Children · Course · FreeFree online safety videos, games and tip sheets for kids from early elementary through high school, covering gaming, generative AI, sextortion and social media. Includes the NetSmartzKids cartoons for younger children. Downloads need free registration.