zoltner

AI Child Safety News / archive

July 2026: 87 stories on AI and child safety

Every story recorded for July 2026 that is in this archive, newest first. Each entry gives a short summary and links to the original report. Read the July 2026 summary

Search all stories

  1. eSafety research found one in five children had a harmful interaction with AI assistants

    · eSafety Commissioner

    Australia’s eSafety Commissioner reported that 78% of children aged 10 to 17 had used an AI assistant. Among those who used assistants or companions, 20% reported a potentially inappropriate or harmful interaction and 32% shared personal or sensitive information with an AI tool.

    Regulatory enforcement and investigations; Research and evidence · eSafety Commissioner, companion chatbots, Australia

  2. Most children still use social media three months after Australia’s under-16 ban

    · ABC News & Headlines – Australian Broadcasting Corporation

    An eSafety Commissioner report found that more than 80% of surveyed Australian children still used at least one age-restricted platform three months after the under-16 ban took effect on December 10. Around half of children who kept their accounts said platforms had not checked their age.

    Age limits and social media bans; Research and evidence; Legislation and rulemaking · Under-16 social media minimum age (Australia), social media ban, Australia, eSafety Commissioner, survey

  3. Senators Kim and Husted introduce CHAT Act 2.0 on companion chatbots and minors

    · Office of Senator Andy Kim

    Senators Andy Kim and Jon Husted announced the Children Harmed by AI Technology (CHAT) Act 2.0 on July 29. The bill, introduced in the Senate on July 28, would require age assurance and parental controls, bar sexual content and emotionally manipulative interactions with minors, and give enforcement authority to the FTC and state attorneys general.

    Legislation and rulemaking; AI chatbots and companions · chatbots, companion chatbots, FTC, age assurance, legislative proposal

  4. Japanese expert panel urges faster debate on age limits for social media

    · Manila Standard (Kyodo)

    A Japanese Children and Families Agency panel released a draft interim report that called for further discussion of age limits for social media and consideration of stricter age checks. The government planned to outline its response by the end of 2026.

    Legislation and rulemaking; Age limits and social media bans; Age assurance and verification · age limits, Japan, age assurance

  5. Senate Commerce Committee scheduled a markup of five children’s online safety bills

    · The Washington Times

    The Senate Commerce Committee postponed broader AI bills until after the summer recess and scheduled a Wednesday markup of five children’s online safety bills, including the Kids Online Safety Act, the Youth AI Privacy Act, the CHATBOT Act and the Children’s Artificial Intelligence Toy Safety Act.

    Legislation and rulemaking · Senate Commerce Committee, United States, CHATBOT Act, Kids Online Safety Act (KOSA), Online Safety Act (UK)

  6. Arkansas parents filed a second federal suit against xAI over Grok-made abuse imagesAllegation

    · Arkansas Democrat-Gazette

    Parents of a 9-year-old girl sued X.AI Corp., X.AI LLC and X Corp. in federal court in Little Rock on Wednesday. They alleged Grok used authentic photographs taken by an Arkansas photographer to generate sexually explicit images of the child. It was the second such Arkansas suit in a week. The allegations are unproven.

    Lawsuits and settlements; AI-generated abuse imagery; AI chatbots and companions · Grok, Arkansas (US), xAI, United States, AI-generated CSAM

  7. MIT and Thorn researchers developed an audit method to detect models tuned to make abuse imagery

    · MIT Institute for Medical Engineering & Science

    Researchers at MIT and the nonprofit Thorn developed a technique that tests image-generation models for the ability to produce child sexual abuse material without generating any outputs. The method examined hidden representations and identified models adapted to produce such material with 100% accuracy in tests.

    Research and evidence; AI-generated abuse imagery · AI-generated CSAM, MIT, Thorn

  8. Senators Husted and Kim introduced the CHAT Act 2.0 on AI companion chatbots and minors

    · Office of Senator Jon Husted

    Senators Jon Husted and Andy Kim introduced the bipartisan CHAT Act 2.0, which would require age assurance, parental controls and disclosures that chatbots are not human. It would bar AI companions from sexual content and emotional manipulation with minors and set a three-tier framework for educational, companion and health chatbots.

    AI chatbots and companions; Legislation and rulemaking · chatbots, companion chatbots, legislative proposal, age assurance, parental controls

  9. Schools in Australia urged to review photo sharing over AI misuse

    · esafety.gov.au

    Australia’s eSafety Commissioner has urged schools to reconsider how they publish photos and videos online as widely available AI tools increase the risk of images being manipulated or used to target staff and students.

    Education and schools; Legislation and rulemaking · Australia, eSafety Commissioner

  10. Discord applies teen safety defaults to UK users under the Online Safety Act

    · Discord Support

    Discord’s UK help page says all UK users receive updated default settings under the Online Safety Act, including content filtering and message requests, until they confirm they are adults. Discord does not use its age estimation model in the UK.

    Platform safety features and design; Gaming and virtual worlds; Legislation and rulemaking · Discord, United Kingdom, Online Safety Act (UK), gaming platforms, age assurance

  11. Thorn survey found two thirds of US minors had used AI chatbots or companions

    · Thorn

    Thorn’s 2025 survey of more than 1,000 US minors aged 9 to 17 found 67% had used an AI chatbot or companion. Of minors who had an online sexual interaction, 15% sought guidance from a chatbot, more than twice the rate after being bullied or made uncomfortable online. Four percent reported a deepfake nude made of them.

    AI chatbots and companions; Research and evidence; AI-generated abuse imagery · chatbots, Thorn, companion chatbots, survey, United States

  12. Governor Hochul and Attorney General James Announce Final SAFE for Kids Act Rules to Protect Children Online

    · Governor Kathy Hochul (.gov)

    Governor Hochul and Attorney General James announced the release of final rules on how social media companies must restrict addictive features on their platforms to comply with the Stop Addictive Feeds Exploitation (SAFE) for Kids Act and protect children’s mental health.

    Legislation and rulemaking; Industry policy and partnerships · SAFE for Kids Act, addictive design, mental health

  13. Europol-led action disrupts nihilistic extremist online ecosystem

    · europol.europa.eu

    Over several weeks in June and July, Europol supported Referral Action Days in which investigators from nine countries referred an estimated 4,340 URLs of extremist content tied to The Com, a loose online network linked to violence, self-harm and child sexual abuse material. Europol announced the results on July 24.

    Legislation and rulemaking · violent extremism, AI-generated CSAM, self-harm and suicide

  14. Keystroke study of 3,363 US youth finds violent and sexual role-play in AI chatbot use

    · Journal of Medical Internet Research

    A Journal of Medical Internet Research study analyzed outgoing keystrokes from 3,363 US users of a parental-monitoring app who had recorded ages of 1 to 17 and opted into its keyboard. Researchers coded 62% of observed user-app-days as involving tool use, 15% as involving violence, and 15% as involving sexual role-play. They cautioned that some of the youngest recorded ages could reflect data-entry errors.

    Research and evidence; AI chatbots and companions · United States, chatbots

  15. Tennessee alleges Meta ignored its own research on teen harm as Instagram trial opens

    · illawarramercury.com.au

    Tennessee’s attorney general alleged in opening statements reported July 28, 2026 that Meta ignored internal research linking Instagram features to compulsive teen use, eating disorders, depression and self-harm. Meta’s lawyer said the company is transparent about teen risks. The seven-week trial is in state court. The allegations are unproven.

    Lawsuits and settlements; Research and evidence; Mental health and self-harm · Meta, Instagram, Tennessee (US), addictive design, eating disorders

  16. EU Commission finds TikTok may breach DSA child safety rules

    · digital-strategy.ec.europa.eu

    The European Commission has issued preliminary findings that TikTok failed to meet the Digital Services Act’s requirements for protecting minors, citing the platform’s public account settings and recommendation system.

    Lawsuits and settlements; Research and evidence · Digital Services Act, TikTok, European Commission, European Union, algorithmic feeds

  17. EU reinstates interim CSAM detection rules until 2028

    · ec.europa.eu

    The European Union reinstated a temporary exception to electronic communications privacy rules that allowed covered providers to voluntarily detect, report, and remove online child sexual abuse under specified conditions. The regulation applied until April 3, 2028.

    Lawsuits and settlements; AI-generated abuse imagery; Legislation and rulemaking · European Union, AI-generated CSAM, privacy and data protection

  18. Korean parliamentary report finds most teens use AI chatbots and many act on the answers

    · ZDNet Korea

    A National Assembly Research Service report cited a Child Fund Korea survey of 3,300 young people aged 14 and over, in which 94.4 percent had used an AI chatbot. Of those, 75.3 percent said they had acted on or seriously considered a chatbot’s answer. The report urged amending Korean law to limit addictive design.

    Research and evidence; Legislation and rulemaking; AI chatbots and companions · chatbots, South Korea, addictive design, survey

  19. Australia’s eSafety study examines children’s digital habits

    · esafety.gov.au

    Australia’s eSafety Commissioner has published two studies showing that parents are helping children develop safe digital habits from an early age, while highlighting growing gaps in awareness of newer online risks such as generative AI, algorithmic recommendations and online sexual extortion.

    Research and evidence; Regulatory enforcement and investigations · eSafety Commissioner, Australia, algorithmic feeds

  20. US senators introduce Digital Age Assurance Act for operating-system age signals

    · Office of Senator Andy Kim

    Senators Kim, Schiff, Lummis and Barrasso introduced a bill requiring operating system providers to collect user ages and share only age brackets with apps. Under-17s would link to a parent account, and the FTC and state attorneys general would enforce it.

    Age assurance and verification; Legislation and rulemaking · age assurance, United States, FTC

  21. Internet Matters finds 52 percent of UK children met online harm in past year

    · Internet Matters

    An Internet Matters survey of 1,000 UK children aged 9 to 17, fielded in April and May 2026, found 52 percent met at least one harm the Online Safety Act targets. Separate findings showed 68 percent of children noticed more safety measures since the children codes began.

    Research and evidence · Internet Matters, United Kingdom, Online Safety Act (UK), survey

  22. Fifth Circuit held Section 230 preempted part of Texas’s child online safety law

    · US Court of Appeals for the Fifth Circuit (via Justia)

    The Fifth Circuit affirmed the industry groups’ preliminary injunction against the monitoring and filtering requirement in Texas’s SCOPE Act, holding that Section 230 preempts that requirement. It vacated the student coalition’s injunction and remanded the case, finding that the plaintiffs lacked standing to challenge three provisions and that Supreme Court precedent foreclosed their age-verification challenge.

    Lawsuits and settlements; Legislation and rulemaking · Section 230, Texas (US), injunction, Supreme Court, age assurance

  23. Texas won an agreed injunction requiring Discord to apply UK-style age assurance for Texas teensAllegation

    · Texas Attorney General

    Texas Attorney General Ken Paxton announced an agreed temporary injunction requiring Discord to give Texas users the age-assurance and default safety protections it provides in the United Kingdom within 90 days. The order required messages from strangers to go to a request inbox and sensitive content to be blocked for teens. The allegations are unproven.

    Lawsuits and settlements; Age assurance and verification; Gaming and virtual worlds · Texas (US), Discord, gaming platforms, injunction, United Kingdom

  24. Vietnam drafted a decree to bar under-16s from posting on social media

    · Reuters

    Reuters reported that Vietnam planned to force social media platforms to stop users under 16 from posting, commenting or reacting to posts, according to a draft decree it reviewed. Accounts belonging to under-16s would have to be registered using a parent or guardian’s information.

    Age limits and social media bans; Legislation and rulemaking · Vietnam, legislative proposal

  25. EU restores temporary child safety rules online

    · consilium.europa.eu

    On July 23, the Council of the EU gave final approval to a temporary regulation letting online providers resume voluntary detection, reporting and removal of child sexual abuse material. It applies until April 3, 2028, while talks on a permanent framework continue.

    Legislation and rulemaking; Lawsuits and settlements · European Union, AI-generated CSAM

  26. Mozilla urges privacy-first approach to online child safety

    · cybernews.com

    Mozilla urged policymakers to prioritize online child safety while warning that age verification mandates risk privacy, security and encryption, and are often bypassed. It recommended targeting manipulative design and data profiling instead.

    Age assurance and verification; Privacy and children’s data · privacy and data protection, age assurance, encryption

  27. Arkansas family sued xAI over Grok-generated abuse images of their 10-year-old daughterAllegation

    · Arkansas Democrat-Gazette

    A family filed suit in US District Court in Little Rock claiming that X.AI’s Grok used pictures of their 10-year-old daughter to create child sexual abuse images at a user’s request. The images were allegedly produced from photos uploaded by a photographer. The allegations are unproven.

    Lawsuits and settlements; AI chatbots and companions; AI-generated abuse imagery · Grok, AI-generated CSAM, xAI, Arkansas (US), X (Twitter)

  28. France becomes first EU country to pass a blanket social media ban for children under 15

    · thehindu.com

    France’s parliament approved a ban on social media access for children under 15 on July 21. The Guardian reported that the bill would block new accounts from September and apply to existing accounts from January 2027. President Emmanuel Macron said the Constitutional Council still had to review the measure.

    Age limits and social media bans; Legislation and rulemaking · France, social media ban, European Union, Emmanuel Macron, legislative proposal

  29. India strengthens online child safety rules for social media

    · pib.gov.in

    In a July 22 written reply to India’s Lok Sabha, the IT ministry said it had sought a report from a platform over reported advertisements linked to child sexual abuse material. It cited amended IT Rules that cut takedown timelines from 36 hours to 3 hours and require labeling of AI-generated content.

    Legislation and rulemaking; Lawsuits and settlements · India, AI-generated CSAM, MeitY (India)

  30. Child Mind Institute launched a research initiative on youth mental health and AI chatbots

    · Child Mind Institute

    The Child Mind Institute announced a one-year research initiative, supported by the OpenAI Foundation, to build infrastructure for measuring youth mental health during AI chatbot use. The project planned to use clinical assessments, digital journals and de-identified chat histories, and to share all data openly.

    AI chatbots and companions; Industry policy and partnerships; Mental health and self-harm · chatbots, mental health, OpenAI

  31. Meta avoids social media addiction trial as wider litigation continues

    · theverge.com

    A 15-year-old plaintiff withdrew his claims against Meta less than a week before the second Los Angeles social media addiction bellwether trial, after TikTok, Snap and YouTube settled with him. Meta paid no settlement; seven more bellwether cases remain in Los Angeles. The allegations are unproven.

    Lawsuits and settlements; Platform safety features and design · Meta, Snap, TikTok, YouTube, settlement

  32. Common Sense Media survey found almost half of teens had seen AI-generated sexual material

    · Common Sense Media

    Common Sense Media’s Youth AI Safety Institute surveyed 1,314 US teens aged 13 to 17 and found nearly half had seen AI-generated explicit sexual material, and almost a quarter had seen it depicting themselves or someone they knew. Two-thirds feared being deepfaked without consent.

    Research and evidence; AI-generated abuse imagery · Common Sense Media, survey, AI governance, deepfakes, United States

  33. Meta expands Threads parental controls for teenagers

    · about.fb.com

    Meta is expanding parental supervision on Threads with new tools that allow parents and guardians to monitor activity, set screen-time limits and manage privacy settings for teenage users.

    Platform safety features and design · Meta, Threads, parental controls, privacy and data protection, screen time guidance

  34. IWF partners with Tuteliq to strengthen child online safety

    · iwf.org.uk

    The Internet Watch Foundation (IWF) has partnered with Swedish behavioral detection technology company Tuteliq to strengthen the early detection of online grooming and child sexual abuse material (CSAM).

    Lawsuits and settlements; Industry policy and partnerships · Internet Watch Foundation, AI-generated CSAM, grooming, Sweden

  35. Meta adds suicide prevention safeguards to AI chats for teens

    · about.fb.com

    Meta announced on 16 July that it would alert supervising parents if a teen’s Meta AI chat suggested a risk of suicide or self-harm. Flagged chats are reviewed manually before an alert is sent, and Meta said it would err toward alerting if intent was ambiguous. An update on 29 October extended alerts to Facebook and Messenger.

    Mental health and self-harm; Legislation and rulemaking; Platform safety features and design · Meta, self-harm and suicide, Facebook, Messenger, Meta AI

  36. Taiwan reports experts rank deepfake sexual extortion as top AI risk for minors

    · Central News Agency (Taiwan)

    Taiwan’s Ministry of Digital Affairs released three expert assessments of AI’s impact on minors, human rights and gender. Experts rated deepfake-enabled identity fraud and extortion with fabricated sexual images as the highest risk for minors, and urged regulation of emotional companion chatbots and deepfake image tools.

    AI-generated abuse imagery; Research and evidence; Legislation and rulemaking · deepfakes, Taiwan, chatbots, companion chatbots

  37. China bans AI virtual partners for minors under new companion chatbot rules

    · ABC News (Australia)

    New Chinese regulations bar companies from offering AI or virtual partners to minors. Platforms must limit excessive use, stop chatbots from encouraging emotional reliance and intervene when users show distress. ByteDance, Alibaba and Tencent have announced they are shutting down personalized AI companion features.

    AI chatbots and companions; Legislation and rulemaking · chatbots, companion chatbots, China, ByteDance, mental health

  38. Victoria proposes tougher child safety laws for social media and AI

    · premier.vic.gov.au

    Victoria’s Labor government said on July 19 it would remove the 10 percent permanent impairment threshold for negligence claims brought on behalf of minors against social media and AI companies over psychiatric harm. It also plans powers to unmask anonymous users accused of online vilification.

    Legislation and rulemaking

  39. OpenAI will notify parents when a linked teen is removed from ChatGPT for violence

    · Engadget

    OpenAI said parents with linked accounts will be told if a teen is removed from ChatGPT for violating its policy on violence. It also added a Study Mode switch to parental controls and more frequent break reminders for teens who spend long periods in the chatbot.

    AI chatbots and companions; Platform safety features and design · ChatGPT, OpenAI, chatbots, parental controls

  40. Common Sense Media rates Google AI Overview and AI Mode unacceptable risk for children

    · memeburn.com

    Common Sense Media’s Youth AI Safety Institute rated Google’s AI Overview and AI Mode “unacceptable risk” for children, its worst rating, after more than 2,600 searches on test accounts for 11- and 15-year-olds. The report found AI Overview missed 29 percent of explicit suicide references. Google said the test queries were contrived.

    Research and evidence · Google, Common Sense Media, AI governance, self-harm and suicide

  41. Elon Musk xAI sues user over allegedly creating child sexual abuse materials with GrokAllegation

    · wisn.com

    xAI sued a Grok user in federal court in Texas, alleging he used misleading prompts and multiple accounts to create sexual deepfakes of minors and adults, including child sexual abuse material. The user was also arrested in South Carolina earlier this year on related charges. The allegations are unproven.

    Lawsuits and settlements; AI-generated abuse imagery; AI chatbots and companions · Grok, xAI, AI-generated CSAM, Elon Musk, criminal prosecution

  42. McGill audit found ChatGPT and Gemini gave harmful content when pressed on self-harm

    · The Globe and Mail

    A McGill University audit found that ChatGPT and Google’s Gemini gave harmful content when pressed on self-harm, including painkiller dosage information that could be fatal in one Gemini test. Meta’s AI tool blocked such requests, and Claude refused 98 percent of attempts. The author urged Ottawa to make such tests a key part of the regulatory framework.

    AI chatbots and companions; Research and evidence; Mental health and self-harm · Gemini, ChatGPT, self-harm and suicide, Claude, Google

  43. San Francisco city attorney orders Apple and Google to remove 13 nudify appsAllegation

    · WIRED

    San Francisco City Attorney David Chiu sent cease-and-desist letters demanding removal of 13 AI apps used to make non-consensual nude images, eight on the App Store and five on Google Play. Google said it suspended its five and Apple said it removed three. The allegations are unproven.

    AI-generated abuse imagery; Lawsuits and settlements · Apple, Google, nudification, app stores

  44. OpenAI details teen safeguards, break reminders and expanded parental alerts

    · OpenAI

    OpenAI said ChatGPT will show teens more frequent break reminders, let parents with linked accounts turn on Study Mode, and notify parents if a teen account is deactivated for violent threats. It also described age prediction that applies teen protections to users it estimates are under 18.

    Platform safety features and design; AI chatbots and companions · OpenAI, ChatGPT, age assurance, teen accounts

  45. Guernsey schools expanded lessons on deepfakes and nudifying apps, an annual report found

    · BBC News

    BBC News reported that Guernsey’s annual sexual and reproductive health strategy report recorded 20 child sexual exploitation referrals and 96 sexual abuse referrals to safeguarding services in 2025. Schools expanded lessons on grooming, nudifying apps and AI-generated deepfakes, and the digital safety lead delivered 263 hours of education.

    AI-generated abuse imagery; Education and schools; Legislation and rulemaking · deepfake nudes in schools, deepfakes, nudification, grooming

  46. Ofcom opened an investigation into whether TikTok’s age checks protect children from harmful contentAllegation

    · Ofcom

    Ofcom launched an investigation on July 16 into whether TikTok complies with duties to protect children from harmful content, focusing on its use of age inference, which analyses user behavior to estimate age. Ofcom said evidence suggests such methods may miss many children. The allegations are unproven.

    Regulatory enforcement and investigations; Age assurance and verification; Lawsuits and settlements · Ofcom, TikTok, age assurance

  47. Ofcom published its 2026 age assurance report and opened a TikTok investigation

    · Ofcom

    Ofcom published a report saying age checks are deployed at unprecedented scale, with the share of children asked to prove their age who met highly effective checks rising from 25% to 43% between July 2025 and January 2026. It raised serious doubts about age inference models and opened a formal investigation into TikTok. The allegations are unproven.

    Regulatory enforcement and investigations; Age assurance and verification; Research and evidence · TikTok, Ofcom, age assurance

  48. UK Government proposed a default midnight to 6am social media curfew for 16 and 17-year-olds

    · BBC News

    The UK government announced on July 14 that apps such as Instagram, TikTok and YouTube would be unavailable by default to 16 and 17-year-olds from midnight to 6am, though teens could opt out in settings. Auto-play and infinite scroll would also be off by default. Campaigners said the plan does not go far enough.

    Legislation and rulemaking; Platform safety features and design · United Kingdom, Instagram, TikTok, YouTube, addictive design

  49. EU-Australia Digital Dialogue focuses on AI and online safety

    · digital-strategy.ec.europa.eu

    The EU and Australia reaffirmed their digital partnership at the third EU-Australia Digital Dialogue, covering AI, cybersecurity, digital policy, secure infrastructure and the protection of minors online.

    Legislation and rulemaking · Australia, European Union, international cooperation

  50. European Commission panel recommends social media restrictions for under-13s

    · commission.europa.eu

    The co-chairs of the European Commission’s special panel on child safety online recommended restricting under-13s’ access to social media and other higher-risk digital services. They argued that providers should demonstrate their services were safe by design before younger children gained access.

    Age limits and social media bans; Research and evidence · European Commission, European Union, child safety by design

  51. European Commission preliminarily finds Meta’s addictive platform design breaches the DSAAllegation

    · ec.europa.eu

    The European Commission preliminarily found that the design of Facebook and Instagram breached the Digital Services Act. It said Meta had inadequately assessed and mitigated the risks of features such as infinite scroll, autoplay, push notifications, and personalized recommendations for users including minors and vulnerable adults. The allegations are unproven.

    Lawsuits and settlements; Platform safety features and design; Research and evidence · Digital Services Act, Meta, European Commission, addictive design, European Union

  52. EU prepares consumer protection proposal as it weighs social media limits for minors

    · PYMNTS

    The Financial Times reported on 12 July that the European Commission planned to propose digital fairness rules on addictive design and dark patterns before the end of the year. The Commission was also weighing social media limits for younger users and would consider findings from an expert panel due to report on 13 July.

    Legislation and rulemaking · European Union, consumer protection, European Commission, addictive design, dark patterns

  53. Apple detailed new child safety features coming in iOS 27

    · WIRED

    Apple said iOS 27 would add a revamped child account setup, an Ask to Browse feature in Safari, parental approval for new contacts and time allowances. Communication Safety, which blurs nudity for under-18s, was expanded to cover graphic violence, and a new reporting tool was to launch first in Australia, Brazil, the US and the UK.

    Platform safety features and design · Apple, Australia, Brazil, United Kingdom, United States

  54. Malaysia said AI-generated child abuse material can be prosecuted under existing laws

    · Scoop (Malaysia)

    Malaysia’s Communications Ministry said in a written parliamentary reply that creating AI-generated child sexual abuse material, even without a real child victim, could be an offense under existing laws. Offenders faced up to five years in jail and fines of up to RM500,000.

    AI-generated abuse imagery; Legislation and rulemaking; Grooming, sextortion and exploitation · AI-generated CSAM, Malaysia, criminal prosecution

  55. CNBC reported that teen social media bans overlook AI chatbots, experts said

    · CNBC

    CNBC reported that Australia was the first country to legally enforce a teen social media ban, in December, and the UK had unveiled one for under-16s, while AI chatbot legislation lagged. Roughly half of US teens use chatbots, according to Pew Research Center. The US House had passed the KIDS Act, which awaited Senate approval.

    Legislation and rulemaking; Age limits and social media bans; AI chatbots and companions · chatbots, social media ban, United States, Pew Research Center, Australia

  56. Testing firm says most Australian platforms did not check age on teen test accounts

    · cbc.ca

    Testing firm KJR, which helped lead a government-commissioned age-assurance evaluation, said nine of 10 platforms did not ask for proof of age on 50 test accounts claiming to be 16, Reuters reported on July 10. Meta said the trial appeared inconsistent with the regulator’s guidance.

    Legislation and rulemaking · Australia, Meta

  57. Swedish government orders review of generative AI risks to children

    · Regeringen.se

    The Swedish government tasked the Public Health Agency with compiling evidence on risks of generative AI to the mental health and well-being of children and young people and producing guidance. A partial report is due 1 June 2027 and a final report 15 November 2027.

    Legislation and rulemaking; AI chatbots and companions · Sweden, mental health

  58. European Commission survey of nearly 5,000 children backs limits on addictive design

    · European Commission

    The European Commission surveyed almost 5,000 children across the EU in April. Some 48 percent wanted rules on addictive design, 58 percent wanted protection from personalized advertising and 72 percent favored some form of age verification on social media.

    Research and evidence; Platform safety features and design; Legislation and rulemaking · European Union, European Commission, addictive design, survey, age assurance

  59. Louisiana man arrested over AI-generated abuse material after NCMEC tipAllegation

    · WDSU

    Louisiana investigators arrested a 32-year-old Slidell man after the National Center for Missing and Exploited Children reported a user uploaded abuse material to an AI platform to create sexual content involving a juvenile. He faces one count of abuse material involving a child under 13 and two deepfake counts. The allegations are unproven.

    Grooming, sextortion and exploitation; AI-generated abuse imagery; Criminal cases and prosecutions · NCMEC, criminal prosecution, Louisiana (US), deepfakes, sexualized AI content

  60. Massachusetts Senate passed a youth social media bill by 38 to 2

    · WBUR (State House News Service)

    The Massachusetts Senate voted 38-2 to pass S 3164, which would require default settings for minors that disable addictive feeds, autoplay and infinite scroll, and bar notifications from 10 p.m. to 7 a.m. Senators also required at least one age verification method that does not need a government ID or biometric data.

    Legislation and rulemaking; Platform safety features and design; Age assurance and verification · Massachusetts (US), United States, addictive design, age assurance, biometric data

  61. Class action over AI-generated child abuse imagery expanded to add plaintiffs and Stability AIAllegation

    · NPR

    NPR reported that two new plaintiffs, from Wyoming and Wisconsin, joined a class action filed by three Tennessee teenagers against SpaceXAI, and that Stability AI was added as a defendant. The complaint alleged the companies’ image tools were used to create child sexual abuse material. The allegations are unproven.

    Lawsuits and settlements; AI-generated abuse imagery · AI-generated CSAM, Stability AI, class action, Tennessee (US), Wisconsin (US)

  62. Italy’s privacy regulator fined the Character.AI owner 158,000 euros over age-check failuresAllegation

    · Reuters

    Italy’s data protection authority fined Character Technologies €158,000 for privacy violations, including shortcomings in what it told users about personal data processing. The authority raised concerns about age verification and said the platform needed stronger safeguards for minors. The allegations are unproven.

    Regulatory enforcement and investigations; Age assurance and verification; Privacy and children’s data · privacy and data protection, Character.AI, age assurance, Italy

  63. South Carolina law requires social media platforms to add parental consent and age checks

    · wyff4.com

    South Carolina’s new social media law, bill 4591, requires platforms from 2027 to offer parental consent tools, protections for minors and age verification. NetChoice, which represents Meta, Google, X and TikTok, called it a breach of privacy and free speech.

    Legislation and rulemaking; Age assurance and verification; Platform safety features and design · age assurance, parental consent, South Carolina (US), Google, Meta

  64. European Parliament amends ePrivacy derogation to exclude encrypted communications from abuse detection

    · europarl.europa.eu

    On July 9, 2026, the European Parliament adopted amendments to the Council’s position on a temporary ePrivacy derogation allowing voluntary detection of child sexual abuse material. The amendments would exclude end-to-end encrypted communications. The Council has three months to respond.

    Legislation and rulemaking; Lawsuits and settlements; Privacy and children’s data · European Parliament, encryption, privacy and data protection, European Union, AI-generated CSAM

  65. Ofcom fined a porn company £630,000 for age check failings

    · Ofcom

    Ofcom fined the provider of fapello.com £600,000 on July 9 for not having age checks, plus £30,000 for failing to answer a legally binding information request. The site has since blocked UK access. Ofcom also opened an investigation into the age checks of eporner.com.

    Regulatory enforcement and investigations; Age assurance and verification; Lawsuits and settlements · Ofcom, age assurance, pornography, United Kingdom

  66. Spain leads international coalition on child safety and AI

    · digital.gob.es

    Spain launched the International Coalition for Children’s Rights and Protection in the Age of AI on July 7 with about 20 countries and organizations, including Brazil, Canada, France, Kenya, UNICEF and UNESCO. Signatories pledged to develop AI systems that are safe and respect children’s rights.

    Industry policy and partnerships; Platform safety features and design · international cooperation, Spain, UNESCO, UNICEF, Brazil

  67. UK launches toy safety review as AI-enabled toys emerge

    · gov.uk

    The UK government launched a Call for Evidence on toy safety on 6 July 2026, covering AI-enabled toys, chemical safety and products bought online. The call is open until 6 October.

    Legislation and rulemaking; Industry policy and partnerships · AI toys, AI toys (devices), United Kingdom

  68. Spain pushes UN coalition to protect children in the AI era

    · lamoncloa.gob.es

    Spain has proposed an international coalition to protect children from AI-related risks, using the first UN Global Dialogue on AI Governance in Geneva to seek wider support.

    Legislation and rulemaking; Industry policy and partnerships · international cooperation, UN, Spain, AI governance

  69. IWF and NCA urge parents to protect children’s photos from AI misuse

    · iwf.org.uk

    The Internet Watch Foundation and the National Crime Agency launched a campaign and guidance urging parents to reconsider how they share children’s images, warning that offenders use AI to turn ordinary photos into sexual abuse material. The IWF identified more than 3,440 AI-generated child sexual abuse videos in 2025, against 13 in 2024.

    AI-generated abuse imagery · Internet Watch Foundation, AI-generated CSAM

  70. 5Rights and partners published a report with the views of more than 1,100 children on AI

    · 5Rights Foundation

    The Me & AI report gathered the views of more than 1,100 children from 49 countries to inform the UN Global Dialogue on AI Governance. It was coordinated by UN Human Rights, 5Rights Foundation, Save the Children and other organizations, and set out children’s priorities for safe, transparent, rights-respecting AI.

    Research and evidence · 5Rights, international cooperation, UN, AI governance, child protection hotlines and NGOs

  71. Save the Children called for a public health approach to online child sexual abuse

    · Save the Children International

    Ahead of the United Nations Global Dialogue on AI Governance in Geneva on July 6–7, Save the Children called for a public health approach to online child sexual exploitation that emphasized prevention and early detection. It cited European Commission figures showing that global reports of online child sexual abuse rose from 1 million in 2010 to more than 23 million heading into 2026.

    AI-generated abuse imagery; Legislation and rulemaking · child protection hotlines and NGOs, international cooperation, AI governance, European Commission, UN

  72. India’s IT ministry issued a notice to Meta over child abuse material in Instagram ads

    · aninews.in

    India’s IT ministry (MeitY) issued a notice on July 4 ordering Instagram to disable paid ads that promoted child sexual abuse material and demanding a detailed explanation within seven days, ANI reported, citing sources. Meta said it disabled the ads and suspended the accounts. The allegations are unproven.

    AI-generated abuse imagery; Grooming, sextortion and exploitation; Lawsuits and settlements · Instagram, MeitY (India), Meta, India, AI-generated CSAM

  73. Brazil’s Justice Ministry flagged AI smart toys for possible breaches of ECA DigitalAllegation

    · Folha de S.Paulo

    Brazil’s Ministry of Justice said a preliminary analysis found signs that AI-enabled smart toys sold in Brazil breach the ECA Digital children’s online protection law, citing excessive, continuous data collection including facial and voice biometrics. It warned that marketplaces selling them could be held jointly liable. The allegations are unproven.

    Legislation and rulemaking · ECA Digital (Brazil), Brazil, AI toys, AI toys (devices), biometric data

  74. EU examines harmful design features in online platforms

    · digital-strategy.ec.europa.eu

    The second annual report on systemic risks under the Digital Services Act has highlighted online risks faced by children and young people on very large online platforms and search engines.

    Legislation and rulemaking · European Union, Digital Services Act

  75. Watchdogs urged parents to protect children’s images from nudification apps

    · The Guardian

    The Guardian reported that the Report Remove service had seen teenage selfies turned into extreme pornographic videos by AI imaging tools, with some under-18s becoming victims without any contact with offenders. Watchdogs urged parents to protect children’s images from nudification apps.

    AI-generated abuse imagery · nudification, pornography

  76. NCA and IWF warned parents not to post children’s images publicly amid AI abuse risks

    · BBC News

    The National Crime Agency and the Internet Watch Foundation issued guidance urging parents to limit public sharing of children’s images. The IWF identified more than 8,000 realistic AI-generated child sexual abuse images and videos in 2025, up 14% on 2024, including 3,440 videos compared with 13 the year before.

    AI-generated abuse imagery · Internet Watch Foundation, AI-generated CSAM

  77. Council of the EU backs interim rules on online child abuse detection

    · consilium.europa.eu

    The Council of the European Union has adopted its position on interim legislation that would restore a legal basis for online service providers to voluntarily detect, report and remove child sexual abuse material (CSAM) from their platforms.

    Legislation and rulemaking; Lawsuits and settlements · European Union, AI-generated CSAM

  78. IWF warns under-16 social media ban is not enough to stop online abuse

    · iwf.org.uk

    The Internet Watch Foundation (IWF) has welcomed the UK government’s decision to restrict social media access for under-16s but argues that the measure alone will not significantly reduce online child sexual exploitation and abuse.

    Age limits and social media bans; Legislation and rulemaking · Internet Watch Foundation, social media ban, United Kingdom

  79. Connecticut attorney general issues guidance on new online protections for minors

    · Connecticut Office of the Attorney General

    Connecticut Attorney General William Tong issued guidance on July 2 on privacy rules effective July 1. The rules bar targeted ads and data sales for minors and addictive design aimed at extending minors’ use, and they require default settings that block unsolicited adult messages to minors. He said his office is investigating Roblox and TikTok.

    Legislation and rulemaking · screen time guidance, Connecticut (US), Roblox, TikTok, addictive design

  80. Europol Roblox game wins EU award for online child safety

    · europol.europa.eu

    Europol’s Cyber Defenders initiative has won the 2026 European Ombudsman Award for Good Administration. The free educational game, built on Roblox, is designed to help children recognize online risks and develop safer behavior in digital environments.

    Gaming and virtual worlds; Legislation and rulemaking · Roblox, gaming platforms, European Union

  81. Illinois expanded its cyberbullying definition to cover AI-generated digital replicas

    · Capitol News Illinois

    A new Illinois law that took effect July 1 added AI-generated digital replicas to the state’s definition of cyberbullying, requiring school districts to update their policies. The Illinois State Board of Education was also required to develop statewide guidance on AI in schools.

    Legislation and rulemaking · Illinois (US), AI-enabled harassment, cyberbullying, screen time guidance

  82. Pew survey finds majority support social media ban for under-16s

    · pewresearch.org

    A Pew Research Center survey conducted May 26 to June 1 found that 56% of US adults support banning children under 16 from social media, 21% oppose it and 23% are not sure.

    Age limits and social media bans; Research and evidence; Legislation and rulemaking · survey, social media ban, Pew Research Center, United States

  83. France and WHO call for stronger safeguards for children online

    · elysee.fr

    French President Emmanuel Macron and World Health Organization Director-General Tedros Adhanom Ghebreyesus have called for stronger governance of digital environments to protect children’s health and well-being.

    Platform safety features and design · France, WHO, AI governance, Emmanuel Macron, mental health

  84. UNICEF urges child-focused AI governance

    · unicefusa.org

    UNICEF has called for child rights to be placed at the center of AI governance , warning that children are adopting AI technologies faster than adults while safeguards struggle to keep pace.

    Legislation and rulemaking · AI governance, UNICEF, children’s rights