Research

Research Spotlight

The Soft-Porn Funnel: Algorithms, Desensitisation and Escalation

What We Looked At

This review examines whether there is evidence for progressive desensitisation and escalation in young people’s consumption of sexual content online, and the role that algorithmic recommender systems play in driving this process. It draws on UK government reports, academic research and investigative journalism to build a picture of how young users move from incidental exposure to increasingly extreme material.

The review uses the term soft-porn to refer to suggestive or sexualised content that may not be explicitly graphic but normalises sexual imagery; content that is widespread on mainstream social media despite age-restriction policies (Flood and Hamilton, 2003; Hirdman, 2007). Extreme or hardcore pornographic content refers to material depicting explicit sexual acts, as well as illegal content incorporating violence, coercion or assault (Children’s Commissioner, 2023; 2025). The boundary between the two is blurred, and the internet creates significant opportunities for young people to encounter both, often without actively seeking them out.

What the Research Found

Almost all children in the UK are online, and exposure to sexualised content begins early. By age 8–9, 97% of children are online, rising to 100% of those aged 10–17, with children spending around three hours a day on screens on average (Ofcom, 2025). The Children’s Commissioner (2025) reports that first exposure to pornography typically occurs by accident between the ages of 11 and 13, and that pornography among young users has become normalised to the point of desensitisation.

Recommender systems are central to the problem. Social media platforms use algorithmic systems that analyse user behaviour to suggest content likely to generate an emotional response, maximising engagement and time on platform (Smith, 2022). When young users interact with sexualised or suggestive content – even passively – these systems interpret that as interest and serve up more. Research by Global Witness (2025), simulating a 13-year-old account on TikTok with restricted mode enabled, found that the platform’s algorithm automatically directed the account toward pornographic content, and that basic interaction caused escalation. BBC News (Spring, 2024) reached similar conclusions, finding that young male users were recommended unprompted violent and sexual material, and that flagging content as “not interested” did little to stop further suggestions.

Pornography creators actively exploit mainstream platforms. Channels operated by pornographers engage young audiences on platforms like TikTok by posting trending dances, fashion and makeup content, while linking to explicit material in their profiles (Hines, 2025). Pornography companies also use advertisements and bots to seed soft-porn content across social media, with more extreme material only a click away (Pilipets et al., 2024). This makes the soft-porn funnel a deliberate business model, not simply a side effect of platform design (Children’s Commissioner, 2023).

Desensitisation and escalation as a possible pattern. Research on adult pornography use suggests that repeated engagement may reduce sensitivity over time, which can prompt users to seek progressively more extreme content (Ince et al., 2024). Whether a similar dynamic applies to young users is less well established. While the Children’s Commissioner (2025) notes that most young people do not actively seek escalating content, some do, which may reflect developing patterns of compulsive use (Andamon et al., 2025). For those who do not seek it out deliberately, algorithmic recommendation may play a role in progressive exposure, though direct longitudinal evidence for this pathway in adolescents remains limited.

Children are particularly vulnerable. Young people’s still-developing cognitive capacities make them less able to fully understand or critically evaluate what they encounter online, and content produced to appeal to them further increases exposure risk (UCL, 2024; Kemp, 2023). Mainstream platforms – including Snapchat, TikTok and YouTube – are regulated by Ofcom and their own Terms of Service, which prohibit pornographic content; however, sexualised material remains persistently accessible in practice (Global Witness, 2025; Children’s Commissioner, 2025).

The consequences for young people are significant. Extended engagement with sexualised content within algorithmic funnels shapes attitudes and behaviour. The Children’s Commissioner (2025) documents how pornography consumption distorts body image, sexual expectations and respect for women. Investigative reporting has highlighted cases where girls felt pressure to alter their appearance and behaviour to match soft-porn creators in order to fit in socially, and where some teenagers have come to regard pornography as a viable career path (Hines, 2025). These findings indicate that the harms extend well beyond exposure itself, affecting how young people understand relationships, consent and their own self-worth.

What This Means in Practice

The evidence points clearly to the role of platform design – not just individual behaviour – in driving young people’s exposure to escalating sexual content. Several implications follow:

• Platform accountability matters. Recommender systems need meaningful reform. Algorithmic tools that are known to funnel young users toward extreme content should be subject to far stronger regulatory scrutiny and enforcement under the Online Safety Act.
• Parental controls are insufficient on their own. Evidence from Global Witness (2025) shows that even restricted mode settings did not prevent harmful content being served to simulated child accounts. Parents and carers cannot be expected to compensate for platform-level failures.
• Education needs to address the algorithmic environment, not just content. Young people benefit from understanding how recommender systems work as a foundation for more critical and self-protective use of social media, including that their engagement shapes what they are shown next.
• The soft-porn pathway is not incidental: it is built in. Pornography companies deliberately use mainstream platforms as a pipeline to introduce young users to increasingly explicit material. Awareness of this as a structural issue, rather than a matter of individual poor choices, should shape both policy and prevention work.

Where the Research Needs to Go Next

The review identifies a gap in understanding how young people engage with sexualised trends on social media – such as viral dances and fitness content – as a form of social participation, rather than passive consumption. The research also does not yet fully explore how the soft-porn funnel interacts with vulnerability factors such as neurodivergent conditions or socioeconomic disadvantage. Rather than extrapolating from adult populations, longitudinal research tracking escalation patterns in young users specifically is a clear priority.

References