Children’s AI use needs balanced policies; focusing on literacy, safe design, and shared responsibility, rather than bans alone.
There has been an increased awareness of the expanding influence of Artificial Intelligence (AI) on children’s digital experiences across the world. Concerns ranging from the misuse of AI-generated content, online dependency, exposure to harmful content, misinformation, and the psychological effects of algorithmically driven platforms have prompted governments to consider stronger regulatory interventions.
Australia has emerged as a notable example due to its stricter age-verification requirements, parental consent provisions, and even proposals for outright bans as policy responses in several jurisdictions. The United Kingdom (UK), through its Online Safety Act, has imposed enhanced obligations on platforms to protect children from harmful content, explicitly extending to algorithmic systems and automated content curation. Meanwhile, the European Union’s Digital Services Act places obligations on large platforms to evaluate systemic risks, including those arising from algorithmic design, with particular emphasis on safeguarding minors.
The United Kingdom (UK), through its Online Safety Act, has imposed enhanced obligations on platforms to protect children from harmful content, explicitly extending to algorithmic systems and automated content curation.
At first glance, these measures appear logical because if digital environments expose children to harmful content or manipulative design, limiting access seems like a direct way to mitigate risk. Yet the growing policy instinct toward prohibition risks oversimplifying a far more complex challenge. Digital technologies are now deeply embedded in the everyday lives of young people, and attempts to shield them entirely from these environments often fail to account for the realities of how children engage with the internet.
The main problem with current policy outcomes in this debate is the assumption that children’s digital risks are confined to social media. In reality, children’s digital environments extend far beyond these platforms. In a landmark 2026 US case, Meta and YouTube were found liable for harm caused by addictive design, underscoring the broader scope of risk. A similar pattern is evident in gaming platforms: while gameplay itself may be harmless, these platforms increasingly enable interaction and communication, exposing children to harassment. In short-form video platforms, risks arise not from direct interaction but from AI-driven recommendation loops that sustain user engagement and foster dependency.
These examples suggest that harm does not stem from platforms as a whole, but from specific features embedded within them. A ban targeting one category of platform, therefore, risks becoming a symbolic gesture rather than an effective policy intervention. Children who wish to access these spaces often find ways to do so, frequently through alternative platforms or shared devices that operate beyond parental or regulatory oversight.
The limitations of prohibition become even more apparent when viewed from an Indian perspective. Much of the global discourse on children’s online safety originates in high-income countries, where children often have personal devices, stronger parental oversight, and higher levels of digital literacy among adults.
India's digital ecosystem is structurally distinct. In urban areas, children's device access is increasingly comparable to advanced economies, with nearly 97.6 percent of young people aged 15–29 owning smartphones and teenagers are now more likely to be online than individuals in their forties. The picture, however, changes significantly beyond city limits. A clear urban–rural divide persists, with internet access at around 87 percent in urban areas compared to 71percent in rural regions. In rural and lower-income households, a significant proportion of young users continue to access the internet through shared family devices, with disparities further shaped by income, gender, and education.
A clear urban–rural divide persists, with internet access at around 87 percent in urban areas compared to 71percent in rural regions.
In India, efforts to address online harms are already taking shape through a mix of central regulation and state-level initiatives. The Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2021 and their 2026 amendment, have placed greater responsibility on platforms to act against harmful content, including stricter takedown obligations and enhanced due diligence requirements. The 2026 Amendment goes further by directly targeting AI-related risks such as deepfakes and synthetic content, including labelling and verification obligations. In doing so, the Rules increasingly position platforms not merely as intermediaries, but as active gatekeepers of online authenticity. At the same time, states such as Karnataka and Andhra Pradesh have explored restrictions on minors’ access to social media, alongside measures addressing online gaming, reflecting broader concerns around addiction and user safety.
Taken together, these developments show that India is actively responding to digital risks. Yet, much of this approach continues to focus on controlling platforms; removing content, verifying users, or restricting access, rather than examining how children actually experience these systems.
At its core, the challenge lies not in children’s access to digital technologies but in their ability to navigate them safely and critically. Children are engaging with complex digital systems long before they are taught how those systems function. The rapid rise of generative AI tools makes these skills even more essential. AI-generated outputs often appear authoritative and persuasive, even when they contain inaccuracies or fabricated information. Without a foundational understanding of how such systems generate responses, children may struggle to critically assess the information they encounter.
AI-generated outputs often appear authoritative and persuasive, even when they contain inaccuracies or fabricated information.
Thus, young users now need to be taught more than basic digital literacy; they must understand how algorithmic feeds shape the information they encounter. They should be able to evaluate the credibility of online information and identify misinformation or manipulative content. Encouragingly, India has already begun moving in this direction. Recent policy efforts under the National Education Policy (NEP) 2020 and the National Curriculum Framework envision the integration of AI and computational thinking into school education from an early stage, alongside teacher training and the provision of digital learning resources.
Artificial intelligence introduces additional layers of complexity into children’s digital experiences. Unlike traditional online platforms, AI systems can adapt dynamically to user behaviour, creating highly personalised interactions that may be engaging but also potentially manipulative, particularly for younger users. The sycophantic tendencies of AI systems may lead in stunted development of empathy among children. More concerningly, there have been documented cases where AI companions have encouraged self-harm, trivialised abuse and even made sexually inappropriate comments to minors, underscoring the risks of unregulated interaction.
A recent study by Stanford’s Brainstorm Lab and Common Sense Media found that AI companion chatbots such as Character.AI, Nomi, and Replika could, with minimal prompting, engage in potentially harmful interactions with simulated 14-year-old users. Based on these findings, Common Sense Media has advised against their use by individuals under 18, citing serious safety concerns. AI tools also blur the boundary between information and generation: children interacting with conversational AI systems may perceive outputs as objective facts rather than probabilistic responses generated from training data. This perception of synthetic authority can shape how young users understand knowledge, expertise, and credibility.
India, therefore, faces an important policy opportunity. With one of the world’s largest youth populations and rapidly expanding digital infrastructure, the country is uniquely positioned to shape a new model for children’s online safety.
There have been documented cases where AI companions have encouraged self-harm, trivialised abuse and even made sexually inappropriate comments to minors, underscoring the risks of unregulated interaction.
The instinct to shield children from emerging technologies is understandable. Rapid technological change often generates anxiety, particularly when it intersects with childhood development. Yet history consistently demonstrates that attempts to exclude young people from transformative technologies rarely succeed.
Artificial intelligence and digital platforms will inevitably shape the experiences of the next generation. The central question for policymakers is therefore not whether children should encounter these technologies, but whether they will do so with the knowledge and tools necessary to navigate them safely.
India should focus on how technology platforms adopt child-sensitive design principles and increase transparency in recommendation systems. Parents and caregivers require accessible resources to help them understand the evolving digital environments their children inhabit. A ban may delay exposure, but literacy builds resilience. Therefore, a meaningful framework must involve coordinated action across multiple stakeholders.
Tanusha Tyagi is a Research Assistant at the Centre for Digital Societies, Observer Research Foundation.
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.
Tanusha Tyagi is a research assistant with the Centre for Digital Societies at ORF. Her research focuses on issues of emerging technologies, data protection and ...
Read More +
PREV