THE AGE OF DATA : ARE WE BECOMING SLAVES TO ALGORITHMS?
Surajit Barua
The invisible intelligence shaping what we see, what we believe, how we live – and what our children learn when no one is watching.
“साविद्यायाविमुक्तये।”
That alone is knowledge which sets us free. — Vishnu Purana
Have you ever wondered how your mobile phone seems to know what you are thinking about?
Imagine a very ordinary situation. One afternoon, you search Google for a particular pair of shoes. You browse a few websites, look at some designs, perhaps compare prices, and then move on with your day. A few hours later, while scrolling through Facebook or Instagram, you suddenly notice advertisements for similar shoes. Open an online shopping platform, and related products may appear there too.
At first, it feels like a coincidence. But when it happens repeatedly, a question naturally comes to mind: how do these platforms know what we are interested in? Is someone watching our activities? Is there something happening behind our mobile screens that we are not aware of?
Consider another experience. You meet a stranger during a journey or at a social gathering. You exchange a few words without sharing names or contact details. Later, that person appears in your Facebook friend suggestions. How is that possible?
Such suggestions may arise from mutual connections, uploaded contact lists, shared networks and other digital signals. The reason behind any one recommendation is usually unknown to us; an unexpected suggestion is not, by itself, evidence that a phone listened to a conversation. Yet these everyday experiences point towards something much larger, something that has quietly become inseparable from modern life: the algorithm.
Having studied computer science, I have always understood that algorithms form the foundation of computing. But what fascinates me today is not merely how algorithms work. It is how deeply they have entered our personal lives — and, increasingly, the lives of our children.
By the Numbers: A Nation Living Online
Before we talk about algorithms, it helps to see how many of us are now online. According to the IAMAI–Kantar Internet in India 2025 study, the country had 958 million active internet users in 2025, nearly 8% more than a year earlier. 57% of them live in rural India, where adoption is growing almost four times faster than in cities [4].
| 958 M active internet users in India, 2025 [4] | 57% of them in rural India [4] | 588 M short-video viewers — 61% of internet users [4] | 44% have used AI features: voice/image search, chatbots, filters [4] |
About 588 million Indians now watch short videos, and 44% of internet users have already engaged with AI-enabled features [4]. DataReportal’s estimates put India’s social media user identities at around 500 million in late 2025, with YouTube’s advertising reach also at about 500 million, Instagram’s at 481 million and Facebook’s at 403 million [5]. These are advertising reach estimates, not exact user counts, but the point is clear enough. Every day, hundreds of millions of Indians are feeding signals into recommendation systems, most of them without thinking about it.

Figure 1. Estimated advertising reach of major platforms in India, late 2025. Ad reach is not equivalent to active users. Source: DataReportal, Digital 2026: India [5].
How the Algorithm Learns About Us
Every search, every video watched, every photograph liked and even the time spent pausing on a post can generate a digital signal. Individually these actions seem insignificant. Analysed together, they can reveal patterns of interests, habits and preferences.
Modern recommendation systems use algorithms, data analytics and artificial intelligence to interpret these patterns and predict what we might want to see, read, watch or buy next. The more we interact with a platform, the more signals it has to personalise our experience. In June 2023, Meta published explanations of 22 AI-powered ranking and recommendation systems across Facebook and Instagram, describing how interactions and predictions decide the order of content [1, 3].

Figure 2. Simplified illustration of personalisation. The signals and ranking methods used vary by platform and are largely proprietary.
How much of what we watch is actually chosen by us? As early as 2018, YouTube’s then chief product officer Neal Mohan said that roughly 70% of the time people spent watching YouTube was driven by its recommendations, rather than by what they searched for [6]. That figure is now old and the system has changed many times since. Still, it tells us something worth remembering: on India’s most popular video platform, the next video we watch is usually something we were offered, not something we looked for.

When Our Choices Become Predictions
Personally, I enjoy watching videos related to geopolitics, international affairs and scientific developments. Whenever I open YouTube, I find recommendations on these subjects. Someone else opening the same platform at the same moment may see cooking tutorials, cricket highlights or comedy shows. The platform is the same; the experience is entirely different.
This can be genuinely useful. We discover educational resources, books, music and products we might otherwise never encounter. But when a system repeatedly shows us what we already like, does it also influence what we will like tomorrow?
Are We All Seeing the Same World?
Consider politics. A person who supports a particular party may naturally engage with speeches, posts and commentary that reinforce that preference. Another person with an opposing view engages with different material. Their feeds can diverge even while they follow the same public events.
A Pew Research Center study published in June 2024 found that among US adults who regularly got news from each platform, 74% of Facebook and 72% of Instagram news consumers said the platform influenced which news stories they saw at least somewhat; the figures were 66% for X and 61% for TikTok [2]. These results measure perceptions, not the proven causal effect of algorithms.

Figure 3. Perceived influence of platforms on news exposure among regular news consumers in the United States. Source: Pew Research Center (2024) [2].
The evidence from India points in a similar direction. In the Reuters Institute’s 2025 India report, YouTube (55%) and WhatsApp (46%) were the leading platforms used for news by the survey sample, followed by Instagram (37%) and Facebook (36%) [10]. The sample is mainly English-speaking and online, so it is not representative of all Indians — but it shows how much of the news diet now passes through algorithmic feeds and forwarded messages.

Figure 4. Platforms used for news in India (past week). Mainly English-speaking online sample; not nationally representative. Source: Reuters Institute Digital News Report 2025 [10].
The largest study of its kind, published in Science in 2023, examined aggregated exposure data covering 208 million US Facebook users during the 2020 election. Ideological separation was greater in what users actually saw than in what was potentially available to them, and greater still in what they engaged with [7]. Yet the story is not simple. Experiments discussed in Nature found that reducing exposure to like-minded content did not necessarily reduce polarisation [8], and a 2025 PNAS study of nearly 9,000 participants found only limited short-term effects of altered recommendations on political opinions [9]. Algorithms shape the information we meet; they do not single-handedly manufacture our beliefs.
Researchers discuss these concerns through the ideas of filter bubbles and echo chambers. Personal choices, social relationships and prior beliefs matter as much as code. Still, the question deserves to be asked: are we freely forming our opinions, or are our opinions being shaped by the information that repeatedly reaches us?
We Were Taught to Protect Our Secrets. What About Our Data?
During childhood, our parents taught us not to share everything with strangers. Some matters belonged within the family; some thoughts were reserved for close friends. Those lessons were not merely about secrecy. They were about protecting personal space.
Today, we share photographs, locations, travel plans, family moments and opinions through our smartphones. We accept permissions and terms of service, often without understanding how information is collected, processed or used. Even when we are not posting, clicks, viewing time and searches provide behavioural signals.
The irony is striking: we were once careful about revealing our personal lives to strangers, yet we now disclose extraordinary details to digital platforms whose data practices we barely understand. This does not mean every application secretly listens to conversations. It means that convenience and privacy require more deliberate choices — from users, and from the law. India has begun to respond: the Digital Personal Data Protection Rules, 2025 were notified on 13 November 2025, with most obligations on data fiduciaries taking effect from 13 May 2027 [21].
When the Algorithm Reaches Our Children
The influence of recommendation algorithms becomes far more serious when we think about children. Globally, UNICEF estimated that roughly one in three internet users is a child [15]. A child may open a video platform to watch a cartoon, a game or a harmless funny clip. But one recommendation leads to another, and sometimes the next video is more shocking, aggressive or disturbing than the last. What begins as entertainment may become repeated exposure to content a child never deliberately searched for.
India’s children are already online
The national ASER 2024 survey of rural India found that 82.2% of children aged 14–16 know how to use a smartphone. Among them, 76% had used it for social media in the week before the survey, compared with 57% for an educational activity [11]. The phone is being used more for scrolling than for studying.
What worried me more was the safety data. Among young social media users, only 62% knew how to block or report a profile, 57.7% how to change a password and 55.2% how to make a profile private — so roughly four in ten could not use even these basic protections. Boys’ awareness of these features was substantially higher than girls’ in most states, and boys were also more likely to own a phone (36.2%) than girls (26.9%) [11].

Figure 5. Smartphone use and online-safety skills among rural Indian adolescents. *Use in the week before the survey. Safety-skill figures refer to social media users. Source: ASER 2024 (Rural), Pratham [11].
Closer home, ASER 2023 found that about 90% of 14–16-year-olds in rural Assam had a smartphone at home, yet only 25.1% of boys and 11.8% of girls in that age group owned one [12]. For most Assamese children, the screen is a shared family device. That matters: on a shared phone, a child may inherit an adult’s recommendation history, and an adult may never see what the algorithm serves the child once the phone is handed over.

What parents are seeing
A LocalCircles survey of urban parents, drawing more than 57,000 responses from 302 districts between November and December 2025, found that two in three parents of school-going children aged 9–17 said their child was addicted to social media, videos or online gaming. Asked how this had affected their children, 61% reported impatience, 58% aggression, 50% hyperactivity and 47% lethargy; only 11% said their children had become happier [13]. These are parents’ self-reported perceptions from an online panel, not clinical diagnoses — but they reflect a widespread anxiety in Indian homes.

Figure 6. Changes urban parents report in children who are heavy users of social media, video or gaming (multiple answers allowed; base 18,523 parents). Source: LocalCircles survey via PTI, December 2025 [13].
This is not only an Indian story. In the United States, Pew Research Center’s 2024 survey of 1,391 teenagers found that 90% use YouTube and nearly three-quarters visit it daily, while 63% use TikTok and 55% Snapchat [14].
What regulators have found
Unlike adults, younger children are still developing the ability to judge context, distinguish exaggeration from reality and recognise manipulation. A violent prank, a fight presented as entertainment or a video glorifying intimidation may appear repeatedly if the system interprets a pause, replay or click as a sign of interest. Yet curiosity, surprise and distress are not the same as genuine approval.
This is not merely a hypothetical concern. Research commissioned by the UK communications regulator Ofcom in 2024 explored how children aged 8–17 encounter violent content online and found that recommender algorithms and group messaging can contribute to exposure. Ofcom has warned that repeated recommendations can narrow what children see and lead them towards increasingly harmful material, and has told technology firms to tame toxic algorithms[16, 17]. The Children’s Commissioner for England reported in 2024 that some children encountered harmful material through recommendations without looking for it, and in 2025 the UK Information Commissioner’s Office described children seeing explicit, violent or disturbing content they did not want to see [18, 19]. A BBC Panorama investigation in 2024 documented similar concerns about violent content being recommended to teenage boys [20].

But we must be careful with the conclusion. Exposure to violent videos does not mean every child will become violent. Effects vary with age, vulnerability, family environment and the nature of the material. The concern is that repeated exposure may normalise aggression, cause distress, encourage imitation of risky behaviour or make harmful material seem ordinary. A direct causal link between any single recommendation and violent behaviour should not be assumed.

The Law Is Catching Up
Governments in several countries have started to treat children’s data, and what children are shown online, as a public concern rather than a purely private matter.
| Where | Measure | What it means for children |
| India | Digital Personal Data Protection Act, 2023; DPDP Rules notified 13 Nov 2025; most duties from 13 May 2027 [21] | A child is anyone under 18. Platforms need verifiable parental consent and must not track, behaviourally monitor or target advertising at children. Breach of child-related obligations can attract penalties of up to ₹200 crore [21]. |
| Australia | Social media minimum age law, in force from 10 Dec 2025 [22] | Under-16s barred from holding accounts on ten major platforms including YouTube, Instagram, TikTok, Snapchat and X. About 4.7 million accounts were deactivated or restricted in the first month; fines up to A$49.5 million [22]. |
| United Kingdom | Online Safety Act; Ofcom children’s safety rules [16] | Platforms told to filter harmful material out of children’s feeds and give children more control over recommendations [16]; the ICO’s Children’s Code covers how children’s data is used [19]. |
| European Union | Digital Services Act, Article 28 [23] | Platforms must ensure a high level of privacy, safety and security for minors, and may not show them advertising based on profiling [23]. |
What makes India’s DPDP law interesting is that it touches the business model itself, which runs on tracking and behavioural profiling. Done well, it could limit what an algorithm is allowed to learn about a 13-year-old in Guwahati or Dhubri. Done badly, with age checks that are easy to get around and consent reduced to a tick-box, it will change very little.
From Morning to Night: A Life Guided by Algorithms
A news app selects headlines; a social platform orders posts; a navigation app suggests a route; a shopping service recommends products; a streaming platform proposes the next film. At every stage, algorithms help organise the choices placed before us.
Of course, we still decide. Algorithms do not literally control every aspect of life. Yet they increasingly shape the information, products and possibilities we encounter before deciding. Perhaps that is their greatest power: they do not need to tell us what to do. Deciding what to show us — and what not to show us — is often enough.
As parents, we ask our children where they are going, whom they are meeting and what they are learning. Perhaps we should begin asking another question with equal seriousness: what is the algorithm teaching them when we are not sitting beside them?
We created algorithms to make life easier. The challenge of artificial intelligence is not only to build smarter machines. It is to ensure that people — and especially children — retain the freedom to think independently, question what they see and make informed choices. A system designed to hold a child’s attention should never be allowed to treat that attention as more important than the child’s wellbeing.
Algorithms may decide what appears on our screens; they must never be allowed to decide who we – or our children – become.
Surajit Barua, AEITO, Assam Pollution Control Board
Mahabahu.com is an Online Magazine with collection of premium Assamese and English articles and posts with cultural base and modern thinking. You can send your articles to editor@mahabahu.com / editor@mahabahoo.com (For Assamese article, Unicode font is necessary) Images from different sources.


















