Skip to content
The Digital Harm Project

A Research Synthesis · 2026

The harms of pornography and online exploitation — and the research, treatment, and tools that counter them.

A synthesis of peer-reviewed research, government statistics, and clinical evidence on exposure, neurobiology, escalation pathways to child sexual abuse material, AI-generated content, treatment effectiveness, and the legislative and technical landscape.

Chapters
9
Citations
261
Reading time
2555 min
Last reviewed
May 2026

Executive Summary

A four-paragraph orientation to the scale, science, and stakes of the issues covered in this research.

Pornography exposure among minors has become a defining public health challenge of the digital era. Research consistently documents that the average age of first exposure now falls between 12 and 13 years, with over half of adolescents encountering explicit content by age 13. The neurobiological evidence is compelling: pornography activates the same mesolimbic dopamine pathways as substance addiction, producing measurable structural brain changes including reduced gray matter in the striatum and weakened connectivity between reward centers and prefrontal cortex. An estimated 3–17% of the population meets criteria for problematic pornography use, with the World Health Organization formally recognizing Compulsive Sexual Behaviour Disorder in the ICD-11.

The relationship between legal pornography consumption and escalation to child sexual abuse material (CSAM) is empirically contested but not negligible. A minority of heavy pornography users — particularly those with pre-existing pedophilic interest, adverse childhood experiences, or compulsive patterns — may escalate to CSAM consumption. Simultaneously, the emergence of AI-generated CSAM has introduced an unprecedented crisis: the Internet Watch Foundation assessed over 8,000 AI-generated CSAM images and videos in 2025, with AI-generated CSAM videos increasing by 26,385% year-over-year. Legislative responses are accelerating globally, including the unanimous Senate passage of the ENFORCE Act in December 2025 equalizing penalties for AI-generated and traditional CSAM.

On the intervention side, Cognitive Behavioral Therapy and Acceptance and Commitment Therapy demonstrate large effect sizes for reducing problematic pornography use, with ACT trials achieving 92% reductions in viewing. Prevention programs from Germany's Project Dunkelfeld to the UK's Inform Plus offer specialized treatment for individuals at risk of CSAM offending, though CSAM recidivism rates remain stubbornly high even among treated populations. Technology solutions span from accountability software and DNS filtering to industrial-scale CSAM detection platforms like Thorn's Safer, which processed over 415 billion files and detected millions of CSAM files in 2025.

Age verification legislation has swept across approximately half of US states following the Supreme Court's June 2025 ruling upholding Texas's law, though research suggests these laws primarily displace traffic rather than reduce consumption. No single intervention — whether technological, therapeutic, legislative, or educational — is sufficient alone. Effective protection requires sustained, coordinated effort across families, schools, technology platforms, legislatures, and clinical settings.

Contents

The nine chapters

  1. Understanding the harm

    How early exposure becomes compulsion, how it escalates toward CSAM, and the AI-generated frontier.

    45–55 min

  2. 01

    Early Pornography Exposure

    Average age of first exposure is now 12–13. Cumulative exposure during adolescence accelerates sharply, and a large share is accidental rather than sought.

    7 min · 3 sections

  3. 02

    Pornography Addiction Pathways

    Internet pornography activates the same reward circuitry as substance addiction. The ICD-11 formally recognizes Compulsive Sexual Behaviour Disorder, with 3–17% of users meeting criteria for problematic use.

    8 min · 4 sections

  4. 03

    Escalation from Pornography to CSAM

    Escalation from legal pornography to CSAM is empirically contested. It is possible but not inevitable, with pre-existing pedophilic interest a stronger predictor than pornography consumption alone. This chapter also covers the production typology behind the imagery — family-circle abuse, self-generated coercion, livestreamed production-on-demand — and the offender-pathway models that map onto it.

    15 min · 5 sections

  5. 04

    AI-Generated Content: The New Frontier

    AI-generated CSAM videos surged 26,385% in 2025. Detection systems built for traditional CSAM cannot identify novel synthetic material. This chapter covers the generation pipeline in practice, the LAION-5B training-data contamination finding, the unsettled US law after Anderegg, the encryption debate, and how to read the headline statistics in light of the Stanford CIS finding that ~80% of "Generative AI" CyberTipline reports involved no AI-generated CSAM at all.

    20 min · 8 sections

  6. The response

    What therapeutic treatment and detection technology actually achieve.

    40–50 min

  7. 05

    Therapeutic Interventions

    CBT and ACT show large effect sizes for problematic pornography use, with one trial achieving a 93% reduction. CSAM-specific programs — Dunkelfeld, Troubled Desire, Stop It Now!, Inform Plus, i-SOTP, CEM-COPE — have varying evidence quality. This chapter also covers survivors of CSAM circulation, whose ongoing-harm experience requires different clinical scaffolding than standard sexual-trauma protocols.

    19 min · 5 sections

  8. 06

    Technology Solutions

    From DNS filters and recovery-program apps to industrial-scale CSAM detection processing 415 billion files annually, the technology landscape spans household tools, structured behavior-change software, hash-sharing networks, distribution-and-economy patterns, the reporting pipeline from CyberTipline to arrest, and the technical detail of how perceptual hashing actually works.

    26 min · 12 sections

  9. Prevention and action

    Prevention strategy, the policy recommendations, and where to get help.

    25–30 min

  10. 07

    Prevention Strategies

    Half of US states now require age verification. The Supreme Court upheld Texas's law in June 2025. Yet evidence shows these laws displace traffic rather than reduce consumption. This chapter also maps how CSAM-specific law differs across the US, UK, EU, Canada, Australia, and the major outliers — the structural variations that shape platform compliance and cross-border investigation.

    13 min · 5 sections

  11. 08

    Recommendations

    Effective protection requires coordinated action across families, schools, technology platforms, legislatures, and clinical settings. No single intervention is sufficient alone. Each section below has a companion guide with the operational details.

    9 min · 8 sections

  12. 09

    Getting Help & Taking Action

    Everything the research points to, gathered in one place: where to report, how to get images removed, where to find help and recovery, the open-source tools, the law, and the full library of role-specific guides. If you need to act, start here.

    6 min · 7 sections

Featured Program

HOPE Program

Established 2008 · California

Outpatient mental health agency offering evidence-based, affordable treatment across California — including a specialized program certified by the California Sex Offender Management Board.

Programs offered

  • Adult Sexual Behavior Program
  • Substance Use & Concurrent Disorders
  • Trauma Program
  • Mood & Anxiety Disorders
  • Domestic Violence Program
  • Youth Forensic Services
  • Anger Management

Clinical approach

  • Cognitive-Behavioral Therapy (CBT)
  • Individual, group, and family counseling
  • Psychological evaluations and risk assessments
  • Relapse prevention planning
  • Specialized tracks for developmentally delayed and dual-diagnosis clients
  • Telehealth and in-person sessions

16 California locations

  • Capitola
  • Concord
  • El Cajon
  • El Cerrito
  • Escondido
  • Hayward
  • Merced
  • Modesto
  • Moreno Valley
  • Salinas
  • San Diego
  • San Francisco
  • San Luis Obispo
  • San Mateo
  • Santa Maria
  • Stockton

The HOPE Program is featured here on its merits as a treatment resource — not a paid placement, and not a sponsor. Its inclusion is editorially independent of the research presented elsewhere in this report. See the About page for editorial policy and the Get Help page for the full list of unaffiliated, free helplines.

If you need help today

You are not alone.

Confidential helplines exist in every category covered in this research — for people struggling with their own use, parents worried about a child, partners affected by a loved one's behavior, and individuals concerned about sexual interest in minors.

See crisis resources →

For families

Layered protection works.

No single tool is sufficient. The strongest defense combines network-level filtering, device-level controls, accountability software where appropriate, and proactive conversation that begins before the average age of first exposure.

Read the technology chapter →

For policymakers

Evidence over intuition.

Age verification laws displace traffic more than they reduce consumption. The Stanford CIS findings on NCMEC data integrity show why inflated statistics misallocate resources. Prevention programs work when they match risk to treatment intensity.

See policy recommendations →