mlsanityData to 14 Oct 2026

Position: Preventing AI-Generated CSAM Necessitates New Approaches to AI Safety

arXiv 2607.05407 ↗First posted 9 Jun 2026cs.CY cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Modern artificial intelligence (AI) systems present profound new risks to child safety. AI is increasingly being misused to create AI-generated child sexual abuse material, facilitate child sexual exploitation, and reduce barriers to harm. In this paper, we argue that protecting children from AI-facilitated sexual abuse requires new approaches to AI safety. Existing safety techniques assume data accessibility, transparency, and evaluation practices that are incompatible with the ethical and legal constraints surrounding child sexual abuse material. We examine how these constraints create new technical challenges, such as limitations on dataset auditing, red teaming, and fine-tuning prevention. In turn, we outline *15 open problems* in online child sexual exploitation and abuse across the AI development lifecycle, from dataset curation and model design to deployment and long-term maintenance. We propose targeted recommendations for researchers, developers, and policymakers to bridge the gap between theoretical AI safety and the realities of child protection. Our work aims to reframe preventing AI-facilitated child sexual abuse as a central, safety-critical dimension for AI research, motivating work that translates responsible AI principles into concrete safeguards against the exploitation of children.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 74 of its 141 references that are in the map. First posted on arXiv 9 Jun 2026.

Builds on

74 of 141 references.

References · 74

  1. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  2. 2021On the Opportunities and Risks of Foundation Modelscited by 4,033
  3. 2014Deep Learning Face Attributes in the Wildcited by 3,693
  4. 2021GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Modelscited by 3,614
  5. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 3,104
  6. 2023Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversationscited by 1,207
  7. 2023Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!cited by 1,154
  8. 2019Machine Unlearningcited by 1,061
  9. 2023Extracting Training Data from Diffusion Modelscited by 732
  10. 2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learnedcited by 729
  11. 2023Towards Automated Circuit Discovery for Mechanistic Interpretabilitycited by 724
  12. 2023Baseline Defenses for Adversarial Attacks Against Aligned Language Modelscited by 651
38 more
  1. 2023Erasing Concepts from Diffusion Modelscited by 539
  2. 2023SmoothLLM: Defending Large Language Models Against Jailbreaking Attackscited by 469
  3. 2023Jailbreak and Guard Aligned Language Models with Only Few In-Context Demonstrationscited by 431
  4. 2022A Holistic Approach to Undesired Content Detection in the Real Worldcited by 356
  5. 2024Deliberative Alignment: Reasoning Enables Safer Language Modelscited by 308
  6. 2024MACE: Mass Concept Erasure in Diffusion Modelscited by 307
  7. 2023The Stable Signature: Rooting Watermarks in Latent Diffusion Modelscited by 304
  8. 2022Red-Teaming the Stable Diffusion Safety Filtercited by 264
  9. 2023An Overview of Catastrophic AI Riskscited by 222
  10. 2020Artificial Fingerprinting for Generative Models: Rooting Deepfake Attribution in Training Datacited by 201
  11. 2026Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnetcited by 191
  12. 2023Tree-Ring Watermarks: Fingerprints for Diffusion Images that are Invisible and Robustcited by 187
  13. 2023Prompting4Debugging: Red-Teaming Text-to-Image Diffusion Models by Finding Problematic Promptscited by 171
  14. 2024Defensive Unlearning with Adversarial Training for Robust Concept Erasure in Diffusion Modelscited by 168
  15. 2023Raising the Cost of Malicious AI-Powered Image Editingcited by 163
  16. 2023Compositional Abilities Emerge Multiplicatively: Exploring Diffusion Models on a Synthetic Taskcited by 93
  17. 2024Unlearning or Obfuscating? Jogging the Memory of Unlearned LLMs via Benign Relearningcited by 84
  18. 2023A Pilot Study of Query-Free Adversarial Attack against Stable Diffusioncited by 76
  19. 2021Can Subnetwork Structure be the Key to Out-of-Distribution Generalization?cited by 75
  20. 2024Jailbreaking Prompt Attack: A Controllable Adversarial Attack against Diffusion Modelscited by 72
  21. 2023SILO Language Models: Isolating Legal Risk In a Nonparametric Datastorecited by 71
  22. 2024On the Societal Impact of Open Foundation Modelscited by 65
  23. 2024Introducing v0.5 of the AI Safety Benchmark from MLCommonscited by 62
  24. 2022Interactive Model Cards: A Human-Centered Approach to Model Documentationcited by 61
  25. 2024Safetywashing: Do AI Safety Benchmarks Actually Measure Safety Progress?cited by 61
  26. 2024R.A.C.E.: Robust Adversarial Concept Erasure for Secure Text-to-Image Diffusion Modelcited by 58
  27. 2023Into the LAIONs Den: Investigating Hate in Multimodal Datasetscited by 54
  28. 2023Spear Phishing With Large Language Modelscited by 52
  29. 2024No Free Lunch in LLM Watermarking: Trade-offs in Watermarking Design Choicescited by 50
  30. 2023Intriguing Properties of Data Attribution on Diffusion Modelscited by 47
  31. 2022Discovering the Hidden Vocabulary of DALLE-2cited by 43
  32. 2023The Journey, Not the Destination: How Data Guides Diffusion Modelscited by 43
  33. 2025Safety Pretraining: Toward the Next Generation of Safe AIcited by 43
  34. 2025Deep Ignorance: Filtering Pretraining Data Builds Tamper-Resistant Safeguards into Open-Weight LLMscited by 42
  35. 2024Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Conceptscited by 40
  36. 2024Adversarial Perturbations Cannot Reliably Protect Artists From Generative AIcited by 38
  37. 2024ART: Automatic Red-teaming for Text-to-Image Models to Protect Benign Userscited by 35
  38. 2025Overcoming Sparsity Artifacts in Crosscoders to Interpret Chat-Tuningcited by 34

50 most cited of 74 All 74 (JSON)

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Beyond "I Can't Help With That": How Child Safety Experts Evaluate AI Chatbot Safety
  2. 2024Exploring Parent-Child Perceptions on Safety in Generative AI: Concerns, Mitigation Strategies, and Design Implications
  3. 2025Safe-Child-LLM: A Developmental Benchmark for Evaluating LLM Safety in Child-LLM Interactions
  4. 2025LLMs and Childhood Safety: Identifying Risks and Proposing a Protection Framework for Safe Child-LLM Interaction
  5. 2024A Safe Harbor for AI Evaluation and Red Teaming
  6. 2025SproutBench: A Benchmark for Safe and Ethical Large Language Models for Youth
  7. 2025Rethinking Data Protection in the (Generative) Artificial Intelligence Era
  8. 2024Toward Safe Evolution of Artificial Intelligence (AI) based Conversational Agents to Support Adolescent Mental and Sexual Health Knowledge Discovery