Child Online Safety

Eleventh Circuit finds Florida's feature-based social media age limits likely constitutional

Eleventh Circuit judges ruled Florida's HB3 minor-account limits likely pass First Amendment review under intermediate scrutiny. The panel found a burden on speech yet called the law content neutral, because it sorts platforms by their features.

Mexico gives Meta, TikTok and X a month to restrict what minors see, with penalties left to a later bill

Sheinbaum signed a decree giving Meta, TikTok, X and other platforms one month to restrict minors' exposure to harmful content. The penalty is a bill the government says it would back, so the first cost is design work against a mid-November date.

Disney hands its first company-wide CTO job to the Character.AI chief who pulled its characters

Karandeep Anand takes over Disney's technology on 2 October, about a year after Disney's lawyers told the company he ran to strip its characters out of user-made chatbots. Several of his engineers come with him.

Brussels moves ChatGPT, Reddit and Roblox into the DSA's strictest tier

The designation converts a growth number into a standing operating cost. Whether that cost or the fine ceiling is the bigger figure depends entirely on how much revenue a company actually has.

The 90% that covered 1.8%: Meta's Take a Break gap hands the states a paper case

An internal document put teen adoption of Instagram's Take a Break at 1.8%. Mosseri's blog post advertised more than 90% retention. Those two numbers are the states' case now.

TikTok ran a placebo safety tool on thousands of users over a staffer's objection, New York alleges

New York alleges TikTok gave thousands of users, including children, a placebo Algo Refresh tool that confirmed a reset it never made. The unsealed complaint also quotes a TikTok well-being manager's written objection, documentary support for the state's claim that the company acted knowingly.

Ofcom investigates whether Meta risk-assessed Instagram's disappearing Instants before launch

Ofcom is investigating whether Meta carried out suitable illegal-content and children's risk assessments before launching Instagram Instants in May. For product teams shipping to UK users, the case turns on when the risk work has to be finished and how much of it counts.

OpenAI previews a college planner for teens as Common Sense Media rates the product an unacceptable risk

OpenAI plans to extend ChatGPT for Teens into college applications and financial aid with a College Planner for U.S. students in grades 10 through 12. It arrives as Common Sense Media's testers say the product's parent alerts failed in extended self-harm conversations, a finding OpenAI disputes.

European Commission's KIDS Act proposal pushes age checks and safe defaults into product design

European Commission's KIDS Act proposal expects services minors can reach to build in age-banded access and safe defaults, with comments due 26 November 2026. Nothing is binding yet. The design it points to ties each default to an age check that keeps no ID documents.

Florida asks a judge to ban ChatGPT's human-like framing and 'conversation prolongation'

Florida AG James Uthmeier asked a court for six orders against OpenAI, including outside approval of new models and a ban on "conversation prolongation." Five of the six govern the shipped product, including minors' access and whether ChatGPT may present itself as human.

California's Adam's Law sets a July 1, 2027 deadline for child safeguards in companion chatbots

Gavin Newsom signed California's SB 1119 on September 10, 2026, setting July 1, 2027 as the start date for its central duties on chatbots used by minors. Companion-chatbot operators with California users have 294 days to build age checks, child-safe defaults and a pre-launch risk assessment into the product.

Florida asks a judge to make outside approval a condition of OpenAI's model development

Florida Attorney General James Uthmeier asked a court on September 28 to bar OpenAI from developing any AI model without independent third-party approval. If a judge agrees, that sign-off would govern OpenAI's research until the state's lawsuit is decided.

Florida asks a court to put OpenAI's next models behind a third party's approval

Florida AG James Uthmeier's 49-page filing asks a state court to bar OpenAI from developing new models until an independent third party approves. OpenAI says it already paused training its top models, so the development fight is over who decides when training resumes.

TikTok agrees to strip the For You feed from Alabama minors in a $100 million settlement

TikTok's Alabama settlement caps underage users at two hours a day

TikTok will pay Alabama at least $100 million and limit underage users to two hours a day in its first state settlement over teen social media addiction. It could owe up to $300 million if the features don't ship, and the list resembles what Meta agreed to in settling with 47 states.

Discord's model sorts most users into teen or adult groups from account behavior

Discord says a behavioral model sorts the vast majority of its users into teen (13-17) or adult (18+) groups without asking for ID. How well that protects teens depends on how often the model mistakes a teen for an adult.

Senators put TikTok's 15 million-user control group on a Sept. 1 clock

Blackburn and Blumenthal cite a sealed 2023 document showing TikTok withheld an anti-echo-chamber safeguard from 10% of US users, and they want a list of every other such test.

Meta's adults-only Muse agent leans on a mascot a child advocate compares to a Teletubby

Meta limits its Muse AI agent to users 18 and over while selling it through Jolly, a dress-up mascot, and a planned Tamagotchi-style device. Fairplay says the character appeals to preschoolers, and Meta answers with extra age checks it has not described.

Meta's settlement writes the teen-safety spec everyone else gets measured against

A judge signed off on the payout, but the enforceable part is a list of product defaults only a parent can undo. Meta's rivals had the terms written for them.

Brussels decided a chatbot's generated answers count as search results

ChatGPT is the first AI assistant pulled into the Digital Services Act. It was filed as a search engine, which puts annual risk assessments, independent audits and researcher data access on OpenAI Ireland by the end of November.