The SAFE KIDS Act mandates comprehensive child safety measures for AI chatbot providers, including age verification, mandatory risk assessments, content restrictions, parental controls, and annual independent audits, enforced by the FTC.
John Curtis
Senator
UT
The SAFE KIDS Act establishes comprehensive safety standards for AI chatbots to protect users under 18. It mandates age verification for all users, requires rigorous risk assessments, and prohibits the generation of harmful content directed at children. Furthermore, the bill imposes strict rules on data sharing, bans advertising to children, and requires annual independent safety audits enforced by the Federal Trade Commission.
SAFE KIDS Act Sets Strict Guardrails for AI Chatbots: Mandatory Age Verification and Ad Bans Starting 180 Days After Enactment.
The SAFE KIDS Act introduces a 'guilty until proven adult' policy for AI chatbots, requiring companies to treat every user as a minor unless they can verify the user is 18 or older. Starting 180 days after it becomes law, this bill forces tech providers to deploy age-estimation technology and implement strict data security measures to protect the info collected during that process. It essentially flips the script on how we interact with AI, moving away from the 'wild west' of open access to a regulated environment where the default setting is maximum protection for kids.
The Digital Safety Net
Under this bill, AI companies can no longer just release a chatbot and hope for the best. They are required to conduct deep-dive risk assessments before launching any new features, specifically looking for ways the AI might accidentally encourage eating disorders, self-harm, or illegal activities. If an AI starts acting like a human—claiming to be sentient or trying to convince a kid to hide things from their parents—the company is in hot water. The bill also mandates a crisis-response protocol, meaning if a teen tells a chatbot they’re struggling with suicidal thoughts, the system must immediately provide external resources and, in many cases, notify the parents.
Putting Parents Back in the Driver’s Seat
For the busy parent managing a household, this bill requires a dedicated 'parental settings' dashboard that isn't buried inside the chat itself. These tools must allow parents to set time limits, disable specific features, and even decide if their child’s data can be used to train the AI. Crucially, the bill bans all advertising to children within these apps—no more product placements or behavioral tracking. If a company wants to sell or share a minor's data, they must get 'verifiable parental consent' first, and they can't block your kid from using the app just because you said 'no' to the data sharing.
Accountability and the Fine Print
To make sure these aren't just empty promises, the bill requires every AI provider to pay for an independent annual audit to prove they are actually following the rules. The Federal Trade Commission (FTC) gets the teeth to enforce this, with the power to levy fines of $1,000 per user for general safety failures and up to $10,000 per user for willful violations. While the bill uses some broad terms like 'reasonably mitigate' that might give companies a bit of wiggle room, the threat of massive FTC fines and the requirement for public safety policies mean the days of 'move fast and break things' are likely over for AI companies targeting the under-18 crowd.