zoltner

Monthly summary

February 2026 summary: AI and child safety

In February the first jury trials against Meta began and the EU preliminarily found TikTok’s design addictive. A New Mexico jury trial opened on 2 February, Mark Zuckerberg testified in a Los Angeles trial on 18 February, and the European Commission accused TikTok of breaching the Digital Services Act with infinite scroll and autoplay. Spain, Poland, Germany and Portugal moved on limits for teenagers, and the Grok investigations widened.

Trials, suits and fines

New Mexico’s attorney general accused Meta of knowingly letting predators use Facebook and Instagram to exploit children; the jury trial began on 2 February. In the Los Angeles trial, Instagram head Adam Mosseri said he did not think users could be clinically addicted and called a plaintiff’s 16-hour day of use problematic, and Zuckerberg said he regretted not identifying users under 13 faster. Meta denied the allegations. The allegations are unproven.

Texas sued Snap on 11 February over its 12+ rating and addictive features, Los Angeles County sued Roblox, and West Virginia sued Apple over iCloud and child sexual abuse material. Egypt blocked Roblox, and the Dutch consumer regulator’s investigation of Roblox continued. Forty attorneys general urged Congress to pass the Senate version of the Kids Online Safety Act. The allegations are unproven.

The UK Information Commissioner’s Office fined Reddit £14.47 million for unlawfully processing the data of children under 13, and Ofcom fined porn companies £800,000 and £1.35 million for having no age checks. The European Commission’s preliminary findings said TikTok’s design encourages compulsive use; TikTok called them false. A federal judge temporarily barred Virginia from enforcing a one-hour default limit for under-16s against NetChoice members. The allegations are unproven.

Grok investigations widened

Indonesia lifted its Grok ban on 1 February under strict supervision. On 3 February Paris prosecutors raided X’s French offices in an inquiry widened to include complicity in possessing child sexual abuse material, and the UK Information Commissioner’s Office opened formal investigations into X and xAI. Ireland’s Data Protection Commission opened its own investigation on 17 February, and Brazil’s data protection authority ordered X to stop sexualized content involving children. The allegations are unproven.

UNICEF, ECPAT and INTERPOL reported that at least 1.2 million children in 11 countries disclosed having images manipulated into sexually explicit deepfakes in a year. More than 100 organizations, including WeProtect, the IWF and NCMEC, launched a “No to Nudify” call for governments to criminalize nudifying tools, and the Sunday Mail reported that CSAM reports on X rose from about 86,000 in 2022 to 686,000 in 2024. The allegations are unproven.

Europe moved on teenagers and platforms

Spanish Prime Minister Pedro Sánchez said Spain will ban social media for under-16s. Poland’s ruling party planned an under-15 ban with fines of up to 6% of turnover, Slovenia prepared draft legislation, Germany’s Social Democrats proposed a ban for under-14s, and Portugal’s parliament approved on first reading a bill requiring parental consent for children aged 13 to 16. The EU reopened the debate on an EU-wide restriction, and the Council of Europe’s human rights commissioner urged regulating platforms rather than children.

The UK government said on 15 February that it will bring all AI chatbot providers under the Online Safety Act’s illegal content duties, and proposed a 48-hour removal rule for non-consensual intimate images with fines of up to 10% of worldwide revenue. Ofcom fast-tracked its decision on hash matching for such images. A UK government study found that 47.8% of parents have ever talked with children aged 8 to 14 about their online experiences.

In Australia, teenagers told the ABC the under-16 ban was not working two months in, and the eSafety Commissioner found large companies were failing to detect child exploitation, including livestreamed abuse. Alaska’s House unanimously passed a bill banning AI-generated child sexual abuse material and limiting minors’ social media, and the UAE banned generative AI use by students under 13 in schools.

Chatbots and evidence

KORA was released as the first public benchmark of how AI models behave with children and teens. A Pew survey found nearly two-thirds of US teens have used AI chatbots and 12% sought emotional support; VodafoneThree found 31% of 11 to 16-year-old chatbot users feel the bot is like a friend; and the UK Safer Internet Centre found 97% of 8 to 17-year-olds use AI tools. Instagram said it would alert parents when teens repeatedly search for suicide or self-harm terms, and OpenAI published a Teen Safety Blueprint for India.

The Washington House passed HB 2225 on companion chatbots 69 to 28, an Oregon Senate committee advanced SB 1546, and an Idaho committee advanced a conversational AI bill. The FTC said it will not enforce COPPA’s parental-consent requirement against data collected solely for age verification. NBC News counted 36 state and federal criminal cases over AI-generated abuse material in 22 states, and Swiss prosecutors investigated a school over AI deepfake nudes of classmates. The allegations are unproven.

What to watch

  • The New Mexico and Los Angeles verdicts.
  • The UK consultation on an under-16 ban, launching in March.
  • TikTok’s response to the Commission’s preliminary findings.
  • Spain’s bill and the French Senate vote on the under-15 ban.
  • Australia’s Age-Restricted Material Codes, due to take full effect on 9 March.
Sources (50)
  1. New Mexico’s child exploitation trial against Meta began with a jury on February 2 The Guardian
  2. Mark Zuckerberg testified in the Los Angeles social media addiction trial against Meta BBC News
  3. European Commission preliminarily found TikTok’s addictive design breaches the Digital Services Act TechCrunch
  4. Under 16 social media ban proposed in Spain euractiv.com
  5. Instagram head said 16 hours of daily use was problematic but not addiction BBC News
  6. Texas sued Snap over its 12+ rating and addictive features Texas Attorney General
  7. Los Angeles County sued Roblox over child endangerment claims County of Los Angeles
  8. West Virginia sued Apple over iCloud and child sexual abuse material Reuters
  9. Egypt blocked Roblox over content harmful to children Anadolu Agency
  10. Dutch watchdog opens Digital Services Act probe into Roblox over child safeguarding euractiv.com
  11. Forty attorneys general urged Congress to pass the Senate version of the Kids Online Safety Act Pennsylvania Office of Attorney General
  12. UK data watchdog fined Reddit £14.47 million over failures to check children’s ages BBC News
  13. Ofcom fined porn company Kick Online Entertainment £800,000 for failing to introduce age checks Ofcom
  14. Ofcom fined porn company 8579 LLC £1.35 million for not having age checks Ofcom
  15. TikTok accused of breaching EU digital safety rules euronews.com
  16. Federal judge temporarily bars Virginia from enforcing SB 854 against NetChoice members Reuters
  17. Grok returns to Indonesia as X agrees to tightened oversight aa.com.tr
  18. French authorities raided X’s offices and the UK opened a fresh investigation into Grok BBC News
  19. UK data regulator opens formal investigations into X and xAI over Grok Information Commissioner’s Office
  20. Ireland’s DPC opens data privacy probe into X’s Grok bloomberg.com
  21. Brazilian regulators ordered X to curb Grok sexualized images of children Leonardi Advogados
  22. UNICEF said 1.2 million children had images turned into sexual deepfakes UNICEF
  23. WeProtect, INHOPE, IWF, NCMEC and 100 groups launched a ’No to Nudify’ call to criminalize nudifying tools WeProtect Global Alliance (No to Nudify call to action)
  24. Childlight says Musk should apologize after reported 697% rise in child abuse material on X dailyrecord.co.uk
  25. Poland’s ruling party plans social media ban for under-15s Reuters
  26. Slovenia plans social media ban for children under 15 cybernews.com
  27. Social media ban for children gains momentum in Germany dw.com
  28. Portugal’s parliament approved parental consent rules for teens on social media Reuters
  29. EU reopens debate on social media age restrictions for children euractiv.com
  30. Council of Europe commissioner urged regulating platforms, not children, over social media bans Council of Europe
  31. UK government will bring all AI chatbot providers under Online Safety Act illegal content duties GOV.UK
  32. UK proposes 48-hour removal rule for nonconsensual intimate images, with fines or blocking Financial Times
  33. Ofcom fast-tracked its decision on hash matching to block illegal intimate images including deepfakes Ofcom
  34. UK government study found fewer than half of parents have talked with children about online experiences GOV.UK
  35. ABC reports Australian teens say the under-16 social media ban is not working two months in ABC News (Australia)
  36. eSafety said major tech firms were failing to detect child exploitation ABC News Australia
  37. Alaska House unanimously passed a bill limiting minors’ social media and AI child abuse imagery Anchorage Daily News
  38. UAE banned generative AI use by students under 13 in schools Gulf News
  39. KORA released as the first public benchmark for AI child safety KORA Benchmark
  40. Pew survey found nearly two-thirds of US teens have used AI chatbots and 12% sought emotional support Pew Research Center
  41. VodafoneThree research found almost a third of children who use AI chatbots see them as friends VodafoneThree
  42. Survey found 97% of UK children use AI tools and 60% of teens fear AI image abuse neowin.net
  43. Instagram will alert parents when teens repeatedly search for suicide or self-harm terms Reuters
  44. OpenAI released a Teen Safety Blueprint for India The Times of India
  45. Washington House passed a bill on AI companion chatbots and minors Washington House Democrats
  46. Oregon Senate committee advanced an AI chatbot safety bill for minors Oregon Capital Chronicle
  47. Idaho Senate committee advances AI safety bill aimed at protecting minors online Idaho News 6
  48. FTC signals flexibility on COPPA age checks iapp.org
  49. NBC News found 36 US court cases over AI-generated abuse material in 22 states NBC News
  50. Swiss prosecutors investigated a school over AI deepfake nude images of classmates World Radio Switzerland

Every story recorded for February 2026 is in the news archive. You can also search all stories.

John Zoltner is the founder and CEO of AIChildSafety.org. About John