U.S. Senate Advances KOSA and New Children’s AI Safety Bills 2026
Children’s privacy is increasingly becoming inseparable from AI governance. On 5 August 2026, the U.S. Senate Committee on Commerce, Science and Transportation advanced the Kids Online Safety Act (KOSA) alongside several bills aimed specifically at protecting children who interact with artificial intelligence. The measures will now move forward for consideration by the full Senate.
The package demonstrates how U.S. lawmakers are expanding the discussion around children’s privacy beyond traditional questions of data collection. Increasingly, the focus is also on how digital products are designed, how algorithms influence young users and how AI systems remember and personalize conversations.
KOSA Returns to the Legislative Agenda
KOSA is not new. A previous version passed the Senate in 2024 by a substantial 91-3 vote but did not ultimately become law. The latest version again seeks to establish a duty of care for covered online platforms and introduce stronger safeguards for minors.
Among its proposed requirements are protective default settings and restrictions relating to potentially addictive design features. The aim is to place more responsibility on platforms for foreseeable harms associated with the way their products are designed and operated.
However, KOSA remains controversial. Supporters argue that stronger obligations are necessary to protect children from harmful online experiences. Critics, including the Electronic Frontier Foundation, warn that enforcing age-based protections could encourage platforms to implement age-assurance systems that require users to provide additional sensitive information.
This highlights an important privacy challenge: measures intended to increase safety can themselves create new privacy risks if their implementation requires greater identification or data collection.
AI Chatbots Come Under Scrutiny
Perhaps the most significant development is the Senate committee’s attention to AI systems used by children.
Three AI-related proposals were advanced alongside KOSA: the Youth AI Privacy Act, the CHATBOT Act and the Children’s Artificial Intelligence Toy Safety Act.
The CHATBOT Act would introduce requirements including family accounts and verifiable parental consent for minors using AI chatbot services. Proposed safeguards also address parental controls, manipulative design techniques and targeted advertising.
The Youth AI Privacy Act focuses in part on how long AI systems should retain and use information from conversations with minors. During the committee process, lawmakers debated whether chatbot memory should automatically be deleted after a set period or whether parents should have greater control over retention choices.
AI-enabled toys are receiving attention as well. Earlier in 2026, senators raised concerns that toys incorporating AI chatbots could collect sensitive information or expose young children to inappropriate responses.
Why Privacy Professionals Should Pay Attention
For privacy professionals and those studying topics covered by certifications such as CIPP/US, CIPP/E and CIPM, these developments illustrate an important direction in privacy regulation.
Children’s privacy is moving beyond consent forms and privacy notices. Regulators and legislators are increasingly considering data minimization, retention limits, profiling, age assurance, protective defaults, product design and AI governance as interconnected issues.
The bills have not yet become federal law, and their wording may change as they move through Congress. Nevertheless, the direction is clear: organizations developing digital products for children may increasingly be expected to demonstrate not only that they process children’s data lawfully, but that their products are designed with children’s privacy and safety in mind from the outset.
For privacy professionals, that makes the intersection of children’s privacy, AI governance and privacy-by-design an area worth watching closely.
