Platform Scorecard

Independent, evidence-based safety ratings for the platforms kids actually use: social media, messaging, gaming, and AI chatbots. The Scorecard gives parents, lawmakers, and advocates one consistent way to compare how well each platform protects children.

What it looks at

Each platform is rated on 43 safety criteria, broken down into 90 finer-grained sub-criteria, across nine areas:

• Identity & age. Whether a child can end up in the wrong age experience, and whether the platform notices.
• Parental tools. How parents link accounts, see activity, set time limits, get alerts, and approve contacts.
• Content safety. What children are shown, and how harmful or exploitative material is filtered and caught.
• Contact safety. Who can reach a child, and how the platform handles grooming, bullying, and sextortion.
• Wellbeing & design. Compulsive-use patterns, social comparison, and how the platform responds in a crisis.
• Data & privacy. How children's data is collected, shared, kept, and used.
• AI companions. Guardrails, emotional attachment, and how AI personas behave with young users.
• Spending. In-app purchases, loot boxes, and whether spending is clear and controllable.
• Accountability. Reporting tools, safety testing, and openness to independent oversight.

How ratings work

• Green The protection exists, works, and is on by default for kids.
• Yellow It exists and works, but a parent has to turn it on.
• Orange It exists but has real limits: incomplete, partly effective, or applied inconsistently.
• Red It's missing, doesn't work, or a child can easily get around it.

Ratings aren't one-size-fits-all. Each one reflects a child's real age, the age the platform thinks they are, and where they live, because a 12-year-old who signed up as an adult has a very different experience from one the platform knows is 12.

How we research it

Every rating comes with a written rationale and cited evidence. We start with what platforms publish (help centers, terms of service, app-store listings, and safety reports) and check it against what safety organizations, regulators, and investigative reporters have found. Research is AI-assisted and reviewed by people, and every finding can be traced to its source. Platforms are re-scored regularly as features and laws change.

Only protections that are live and working count. Platforms get no credit for features that are announced, planned, or in beta. The Scorecard doesn't label platforms simply “safe” or “unsafe,” and it doesn't rate on reputation or headlines. It shows exactly where each platform protects kids and where it doesn't.