Childhood in the Age of Algorithms: Lessons from 2025
पुस १५, २०८२ १६:३९
As 2025 draws to a close, one reality is increasingly hard to ignore: childhood is being reshaped by digital technologies far faster than systems of protection, governance, and accountability are evolving. For today’s children and young people, digital spaces are no longer optional or peripheral. They are classrooms, social arenas, creative outlets, and sites of identity formation. Yet, these environments remain largely designed for adults and commercial interests, rather than for children’s rights, safety, or well-being.
This imbalance has significantly shaped 2025.
Digital technologies have undoubtedly created opportunities. Children are learning online, expressing themselves creatively, and connecting across borders in ways unimaginable a generation ago. AI-supported tools have helped some overcome barriers linked to geography, disability, or limited resources. Online communities have also provided spaces of belonging, particularly for those who feel isolated or marginalised offline.
But these benefits sit within a digital ecosystem driven by scale, engagement, and data extraction. Globally, one in three internet users is a child. Children are entering digital spaces earlier, spending more time within them, and leaving increasingly detailed digital footprints. Algorithm-driven systems shape what they see, who they interact with, and how they understand themselves. In this environment, children are not only users of technology; they are also data sources, targets, and increasingly, commodities.
AI and Children
One of the most consequential developments of 2025 has been the rapid integration of generative AI into everyday digital life. AI systems now produce text, images, videos, and personalised recommendations at scale. Increasingly, they also interact directly with children through conversational agents and AI chat companions, often marketed as friendly, supportive, and emotionally responsive.
For children and adolescents, this raises serious concerns. AI chat companions blur the line between human relationships and machine interaction at a critical stage of social and emotional development. They are not neutral. They reflect the values, biases, and commercial priorities embedded in their design, while operating with limited transparency and accountability. As highlighted in the WeProtect Global Alliance’s Global Threat Assessment 2025, emerging technologies are not simply introducing new risks; they are amplifying existing harms, enabling abuse to occur at greater speed, scale, and sophistication, while making detection and accountability more difficult.
This amplification is evident in the evolving landscape of online violence against children and young people. Long-standing risks such as grooming and sexual exploitation have not diminished. They have adapted.
Global data underline the scale of this shift. The 2025 mid-year CyberTipline report by the US-based National Center for Missing and Exploited Children shows a sharp rise in online crimes against children, driven largely by AI-enabled abuse and financial sextortion. Reports of AI-generated child sexual exploitation increased from 6,835 in the first half of 2024 to more than 440,000 in the first half of 2025. Online enticement reports rose by 77 per cent, while financial sextortion cases reached nearly 23,600, often targeting teenage boys. Reports of child sex trafficking also increased alongside growing concern about violent online groups coercing children into self-harm. These figures represent reports, not individual victims, suggesting the true scale of harm is far greater.
Image-based abuse has become particularly pervasive. Childlight – Global Child Safety Institute’s global report Into the Light estimates that millions of children worldwide are affected by image-based sexual abuse, much of it hidden, underreported, and unresolved. Deepfake technology has intensified this harm by enabling the creation of sexualised images without any original material, undermining consent and making redress extremely difficult.
These global trends are mirrored in Nepal. Data from the Nepal Police Cyber Bureau show that reported cybercrime has increased nearly fivefold in four years, from 3,906 cases in 2020/21 to 18,926 cases in 2024/25, averaging more than 50 cases a day. The nature of these crimes is also changing. Reports linked to TikTok have risen by around 3,000 per cent, demonstrating how rapidly emerging platforms can become sites of harm when safeguards are weak. While Facebook remains the most frequently reported platform, the broader pattern points to an increasingly complex and evolving threat landscape.
Children and young people are disproportionately affected by online sexual abuse, blackmail, image-based exploitation, impersonation, and harassment. Underreporting remains a major challenge, particularly where stigma, fear, or lack of awareness prevent children from seeking help. Even so, the upward trend is a clear warning. As digital access expands, risks are outpacing awareness, safeguards, and response capacity.
Beyond sexual exploitation, technology-facilitated gender-based violence has become more visible. Adolescents, especially girls and marginalised groups, face harassment, threats, and coordinated abuse across platforms. Algorithms that reward outrage amplify harm, while addictive design features and constant comparison through likes and follower counts contribute to anxiety, low self-esteem, and emotional distress.
Social Media Restrictions
Governments are beginning to act. On 10 December 2025, Australia introduced restrictions on social media use for children under 16. However, age-based restrictions risk creating a false sense of safety, as children often bypass them or are pushed into less visible, less regulated spaces. Global child rights organisations such as Save the Children and ECPAT International, alongside ChildSafeNet, have cautioned that restrictions alone will not protect children from online harms. More fundamentally, such measures risk framing children as the problem, rather than confronting why digital environments are so unsafe that exclusion appears to be the only solution.
The core issue is not whether children should be online. It is whether digital spaces are designed and governed in ways that respect children’s rights, capacities, and developmental needs.
This became starkly clear in Nepal in September 2025, when the sudden ban on 26 social media platforms triggered a Gen Z–led movement that escalated into widespread unrest. For many adolescents and teens, the internet is not a luxury but a primary channel for learning, expression, and social connection. The blanket restrictions fuelled anger, misinformation, and rapid mobilisation, showing how digitally disconnected policies, introduced without consultation, can destabilise societies and carry serious political consequences.
Safety-by Design
A more effective response requires a fundamental shift. Child safety must be built into digital products by design, with strong privacy defaults, limits on data collection, transparent recommendation systems, and clear boundaries on how AI systems interact with children. Governments must move beyond statements of intent to enforceable laws, meaningful penalties, and independent oversight. Technology companies must be held accountable not only for responding to harm, but for preventing it.
Children’s participation is essential. Their lived experiences must inform policy and product design. Parents and educators need support, not blame. Expecting families to manage complex and rapidly evolving digital risks alone is unrealistic.
The lessons of 2025 are clear. Digital technologies are not inherently harmful or beneficial. Their impact depends on the choices made by governments, technology companies, and society as a whole. continue to be embedded in digital systems.
Children and young people have the same right to protection online as they do offline. Governments, technology companies, institutions, and communities must act with urgency and responsibility. The digital world is already shaping childhood. The question is whether we will shape it in ways that truly protect children.
(Author Raghuvanshi is the founder and president of ChildSafeNet, an organization dedicated to the welfare and protection of children in the digital era. He is a child protection activist.)
पछिल्लो अध्यावधिक: पुस १५, २०८२ १६:५२
