Growing up in today’s world involves an increasing amount of interaction with technology. The rise in availability, accessibility, and use of the internet, along with social norms that encourage internet connection, make it nearly impossible for children to avoid online engagement. The internet undoubtedly benefits children socially and academically and mastering technological tools at a young age is indispensable for opening doors to valuable opportunities. However, the internet is risky for children in myriad ways. Parents and lawmakers are especially concerned with the tension between important advantages and risks technology bestows on children.

New technological developments in artificial intelligence are beginning to alter the ways parents might choose to safeguard their children from online risks. Recently, emerging AI-based devices and services can automatically detect when a child’s online behavior indicates that their well-being might be compromised or when they are engaging in inappropriate online communication. This technology can notify parents or immediately block harmful content in extreme cases. Referred to as algorithmic parenting in this Article, this new form of parental control has the potential to cheaply and effectively protect children against digital harms. If designed properly, algorithmic parenting would also ensure children’s liberties by neither excessively infringing their privacy nor limiting their freedom of speech and access to information.

This Article offers a balanced solution to the parenting dilemma that allows parents and children to maintain a relationship grounded in trust and respect, while simultaneously providing a safety net in extreme cases of risk. In doing so, it addresses the following questions: What laws should govern platforms with respect to algorithms and data aggregation? Who, if anyone, should be liable when risky behavior goes undetected? Perhaps most fundamentally, relative to the physical world, do parents have a duty to protect their children from online harm? Finally, assuming that algorithmic parenting is a beneficial measure for protecting children from online risks, should legislators and policymakers use laws and regulations to encourage or even mandate the use of such algorithms to protect children? This Article offers a taxonomy of current online threats to children, an examination of the potential shift toward algorithmic parenting, and a regulatory toolkit to guide policymakers in making such a transition.