Social media algorithms don't simply show us posts from friends and creators we follow.
They also recommend content based on signals such as our interests, behaviour and activity on the platform.
That means what appears in our feeds isn't always content we've chosen to see.
And this raises an important question: what happens when the content that attracts the most attention is also harmful?
For young people, understanding how social media recommendation systems work is becoming an essential part of digital literacy.
How do social media algorithms work?
When we open Instagram, TikTok or another social media platform, we're not necessarily seeing content in the order it was posted.
Instead, recommendation systems help determine what appears in our feeds.
These systems can use information about our activity, including the videos we watch, posts we interact with and accounts we follow, to predict what content might interest us.
The more we interact with particular types of content, the more the platform may recommend similar material.
This can be useful.
It helps people discover new creators, learn about their interests and find content they enjoy.
But it also means platforms can influence what we see, even when we haven't actively searched for it.
And that's something young people need to understand.
Why can social media algorithms recommend harmful content?
One of the concerns with recommendation systems is that content attracting strong reactions can also generate significant engagement.
And engagement is valuable to social media platforms.
The longer people spend watching, scrolling and interacting, the more opportunities platforms have to display advertising and collect information about user behaviour.
This creates a commercial incentive to keep people engaged.
The problem is that some content attracting attention can also be harmful.
Recommendation systems can expose users to misogynistic, sexist, violent or increasingly extreme material.
Young people don't necessarily need to search for this content.
It can appear through recommendations based on signals the platform has collected about their activity.
This doesn't mean every recommendation system deliberately promotes harmful material or that every user will experience the same thing.
But it does mean we need to pay attention to how these systems operate and the risks they can create.
Why repeated recommendations matter
Imagine a young person watching a few videos about fitness.
The platform begins recommending more fitness content.
Over time, those recommendations might become increasingly focused on unrealistic expectations, appearance or extreme exercise advice.
Another young person might watch motivational videos and gradually encounter content promoting hostile attitudes towards women.
These examples show how recommendation systems can potentially introduce people to material they never intended to seek out.
The concern isn't only about one harmful video appearing in someone's feed.
It's about repeated exposure.
When young people see particular messages again and again, those ideas can begin to feel more common or widely accepted than they actually are.
A social media feed is not necessarily an accurate reflection of what most people think or believe.
It's a collection of content selected by a recommendation system.
And those systems are influenced by what platforms predict will keep users engaged.
What is algorithmic literacy?
Algorithmic literacy is the ability to understand how algorithms influence the information and content we encounter online.
For students, this means recognising that their social media feeds aren't simply a collection of posts from people they follow.
They are shaped by recommendation systems making decisions about what to show them.
Algorithmic literacy also involves understanding why certain content might be recommended and how those recommendations can influence our perceptions.
Students should be able to ask:
Why am I seeing this content?
Did I choose to follow this account, or was the post recommended to me?
What information might the platform be using to make this recommendation?
Am I seeing different perspectives, or mostly similar opinions?
Is this content accurate, trustworthy or designed to provoke a reaction?
These questions help young people become more critical consumers of technology.
They also create opportunities to discuss how online platforms make money and why keeping people engaged is such an important part of their business models.
Australia's Digital Duty of Care and algorithmic choice
Australia's proposed Digital Duty of Care could give people more control over what appears in their social media feeds.
Under the Government's My Feed, My Way initiative, Australians over 16 would be able to choose between algorithmically recommended content and a feed showing posts from accounts they follow.
The proposal forms part of the Online Safety Amendment (Digital Duty of Care) Bill 2026, introduced to Parliament in October 2026.
If implemented, this could give users more control over the content they encounter.
Instead of relying on platforms to select recommendations based on their behaviour, people could choose a following-only feed.
This would be a significant change to how many people currently experience social media.
However, switching off personalised recommendations wouldn't automatically remove every online safety risk.
People could still encounter harmful content through accounts they follow, and platforms would still have other features influencing how users interact with content.
The proposal is one part of a much broader conversation about platform accountability and online safety.
Why schools need to teach algorithmic literacy
We spend a lot of time teaching young people how to use technology.
But we also need to teach them how technology influences them.
That includes understanding how social media platforms decide what appears on their screens.
Young people need opportunities to recognise that algorithms are designed systems, not neutral reflections of the world.
They should understand that recommendations can be influenced by their behaviour and that platforms have commercial reasons to keep them engaged.
This is an important part of digital wellbeing and online safety education.
It's also closely connected to critical thinking.
When students understand how recommendations work, they're better equipped to question the content they encounter rather than simply accepting it.
They can begin to recognise when their feeds are presenting a narrow range of opinions, when content is designed to attract attention and when they need to look beyond what the algorithm is showing them.
A simple classroom activity to explain social media algorithms
Teachers can introduce algorithmic literacy without needing to explain complex computer science.
Start by asking students to think about how content appears in a social media feed.
Discuss the difference between content from accounts they deliberately follow and content recommended by a platform.
Then explore a few questions together:
1. What might influence a recommendation?
Discuss how watching, liking, sharing and interacting with posts can provide signals to a platform.
2. Why might platforms recommend content that attracts strong reactions?
Explore the relationship between engagement, attention and advertising.
3. Does seeing something repeatedly mean it's widely believed or accurate?
Discuss how recommendation systems can create a misleading impression of what is normal or popular.
4. What choices can users make?
Explore how people can manage recommendations, review who they follow and question the content appearing in their feeds.
The aim isn't to convince students that every algorithm is harmful.
It's to help them understand that these systems influence their online experiences and that they should be able to question how they work.
Young people need to understand who's choosing what they see
Social media algorithms play a significant role in shaping the content we encounter every day.
They can help us discover useful information and connect with our interests.
But they can also introduce us to content we haven't searched for, reinforce particular viewpoints and expose users to harmful material.
That's why algorithmic literacy matters.
Young people need to understand that their feeds are shaped by recommendation systems, not simply by the accounts they have chosen to follow.
They also need to understand the commercial incentives behind those systems and how repeated exposure to particular content can influence their perceptions.
Through my Truth or Algorithm and Dopamine Hackers workshops, I help students explore how digital platforms influence what they see, think and do online.
Because if we want young people to make informed choices about technology, we need to help them understand who's choosing what they see, and why.
