The relationship script now arrives by feed

I keep coming back to how late school often arrives to this conversation. By the time adults wheel out the laminated consent lesson, plenty of young people have already met the real script somewhere else: in a group chat, on TikTok, in porn, in the algorithmic sermon that tells them what a girl owes and what a boy can ask for. A new Our Watch and University of Melbourne report puts numbers to that uneasy feeling. Romantic relationships can begin around ages 10 to 13. Sexual activity commonly starts around 15. The script is arriving early.

What Patty Kinnersly from Our Watch is really arguing is that prevention has been mis-timed. We still talk as if violence begins in adulthood, when the attitudes underneath it are already being rehearsed in bedrooms, schoolyards and private message threads. The relationship script is now being drafted before the first date, before the first kiss, before a lot of parents would say the subject has even started.

But Chanel Contos, in Elle Australia’s conversation about the manosphere and algorithms reads the same problem with a harder edge. Even if schools improve, the feed does not clock off at 3pm. It keeps teaching. One camp says start earlier. The other says earlier is not enough when the loudest teacher in the room is a platform.

That is why this lands as a relationships story, not a policy recap. It is about the emotional education of growing up online, the way intimacy gets imagined before it is lived. A Conversation analysis on sex and relationships education makes the point plainly: if teenagers do not get usable language for coercion, respect and repair from trusted adults, they will borrow it from wherever they can.

Before the first date

The number that stays with me is 10 to 13. That is not the age when most adults imagine a serious conversation about relationship culture beginning, yet it is the age range the Our Watch report identifies as the point when many young people start forming romantic relationships. The report also says the average age of first exposure to pornography is 13.6 years. So the sequence matters. A child can encounter performance, power and shame online before they have much practice with tenderness, negotiation or the ordinary awkwardness of saying no. By the time a teacher tries to explain consent, the feed may already have supplied the more vivid lesson.

Kinnersly’s question is the right one: how early do we start if the culture is already there? Her answer is not panic but proportion.

Gender-based violence doesn’t start in adulthood. It is shaped by the attitudes, norms and behaviours that young people are exposed to early in life.

— Patty Kinnersly, Our Watch

That line matters because it pulls the conversation away from one awful incident and back towards atmosphere. Atmosphere is harder to measure and easier to miss. It is the joke passed around as flirtation. The insistence that jealousy proves care. The assumption that relentless access is romantic rather than invasive.

Burnet Institute’s latest survey of young Australians helps answer another of Megan Lim’s questions: which relationship skills are getting most distorted. The uncomfortable answer is the quiet ones. Boundaries. Communication. Checking in. Seventy-seven per cent of respondents said sexual pleasure was not covered at school, and only 42 per cent said consent education was covered well. When the formal lesson is thin, the informal one gets muscular.

Porn and social media don’t show the conversations that make relationships healthy, things like boundaries, consent and checking in with each other,

— Megan Lim, Burnet Institute

That is such a clean diagnosis of the problem that it almost reads like a brief against the whole internet. Not because every teenager is being radicalised by a phone, which is too simple, but because online culture is very good at staging the visible parts of desire and terrible at staging the negotiations underneath it.

The lesson the feed keeps teaching

No one in the formal material uses the phrase shadow curriculum, but that is what this looks like from the outside: a second, more seductive programme running alongside school and family, complete with visuals, repetition and status rewards. It is inconsistent, sometimes funny, sometimes ugly. But it is always on.

That is where the Australian Human Rights Commission’s On Your Terms survey sharpens the picture. Deb Tsorbaris is less interested in lecturing young people than in acknowledging a basic fact: if the official lesson is stale, they will go elsewhere.

If we don’t offer relatable and relevant sex and consent education, young people will turn to less reliable sources like porn, social media or peers,

— Deb Tsorbaris, Australian Human Rights Commission

This is the policy perspective in the best sense. It is admitting that adults have left a vacuum and are surprised when the vacuum fills itself.

The sceptic case, though, is harder to shake. The Conversation’s recent analysis of 142 men’s TikTok histories argues that misogyny online is not a fringe eruption but part of a broader masculinity content ecosystem. Read beside Contos’s warning, it suggests consent education is also a distribution problem. The lesson is competing with a machine built to keep attention hot and gender roles theatrically sharp.

That helps explain why the harm described in the Our Watch material does not feel hypothetical. Nearly one in three young people who had had sex reported unwanted sex, some of it before 14. Young women, and trans and non-binary young people, are more exposed to the sharp end of this. BBC reporting on coercive control is useful here because it reminds us how abuse often arrives dressed as ordinary relationship management: access to money, time, your own phone, your own head. The teenage version may look smaller at first. It is not necessarily smaller.

What school can’t outrun

I do not think the answer is to speak about young people online as if they are passive. That lets platforms off too easily and flattens teenagers into victims of technology. They are reading each other, trying on personas, stealing language, discarding it, starting again. Messily. Like everyone else. But the conditions of that experimentation have changed. The Conversation’s argument for criminalising sexualised deepfakes and going further upstream is a reminder that the next version of sexual harm is often technical before it is fully social. Schools can help. They cannot outrun the tools.

So what does a better response look like? Probably less moral panic, and less fantasy that one good lesson in Year 10 will sort it out. Kinnersly’s prevention lens, Lim’s focus on missing skills, Tsorbaris’s insistence on relevance and Contos’s suspicion of the feed are not competing diagnoses so much as layers of the same one. Start earlier, yes. Make it specific. Teach boundaries and repair, not just legal definitions. Treat algorithms as part of the relationship environment, not a side issue. The task is not simply giving young people information. It is giving them sturdier stories about what care, refusal, pleasure and mutuality are meant to feel like.

I might be wrong, but this is where the report lands for me. We have spent years treating digital life and relationship life as adjacent, when for teenagers they are often the same room. The phone is not interrupting the script. It is helping write it.