{
  "slug": "ai-deepfake-intimate-image-adult",
  "question": "What are the odds an AI-generated intimate deepfake of you will be created or shared without consent in your lifetime?",
  "category": "tech",
  "tags": [
    "digital-fraud"
  ],
  "no_reliable_estimate": false,
  "perceived": {
    "description": "Public discourse about non-consensual intimate deepfakes is dominated by a handful of celebrity cases — Taylor Swift in January 2024, women in the US Congress later that year, K-pop idols across multiple incidents. The mental model that emerges is \"this happens to famous women on the internet,\" which lets ordinary adults file the risk under things-that-happen-to-other- people. The cost trajectory of the underlying tools does not support that framing: open-source models that required a research-grade GPU in 2019 now run on consumer phones, and a usable face-swap on a single photograph takes minutes. Asked directly, most adults significantly underestimate the population prevalence of personal victimization, which a peer-reviewed survey of ~16,000 respondents across 10 countries put at 2.2% as of 2023 — in absolute terms, several million people, drawn overwhelmingly from a single half of the population.\n",
    "rough_estimate": "Most adults file this under 'happens to celebrities' rather than a personal risk",
    "kind": "intuition"
  },
  "native": {
    "display": "~2 in 100 (ever, across 10 countries 2023)",
    "numerator": 22,
    "denominator": 1000,
    "unit": "lifetime-to-date (2023)",
    "population": "adults aged 16+ across 10 countries (Australia, Belgium, Denmark, France, Italy, Netherlands, Mexico, South Korea, UK, US), Umbach et al. CHI 2024"
  },
  "normalized": {
    "lifetime_us_adult": 0.04,
    "display": "~1 in 25 over a typical adult lifetime",
    "log_value": -1.4,
    "assumptions": "The 2.2% figure from Umbach et al. (2024) is a contemporaneous \"ever experienced\" prevalence measured in 2023, when consumer-grade deepfake tools had been broadly accessible for roughly five years. Translating it into a remaining-lifetime probability for a US adult requires two adjustments in opposite directions. First, the headline understates true victimization because (a) many victims do not know intimate deepfakes of them exist — the imagery is shared in closed channels or hosted on dedicated sites the subject never visits, and (b) self-report surveys chronically under-capture sexual-violence categories due to recall and disclosure barriers, with Eaton et al.'s (2017) US-only NCII survey finding 8% lifetime prevalence on a broader definition that includes real (non-AI) intimate imagery. Second, the technology curve is steep: the Home Security Heroes census found a 550% increase in detected deepfake videos from 2019 to 2023, almost entirely pornographic and targeting women. A reasonable lifetime estimate combining the peer-reviewed contemporary base rate with a moderate growth assumption over a 59-year horizon lands near 4% — roughly double the 2023 prevalence, well below Eaton's broader-definition 8% which captured fifteen years of pre-AI image-based abuse. The uncertainty band is wide (2%-10%) and skews upward because the technology is still in rapid diffusion.\n",
    "uncertainty": {
      "low": 0.02,
      "high": 0.1
    },
    "scope": "us_adult_lifetime"
  },
  "sources": [
    {
      "url": "https://arxiv.org/abs/2402.01721",
      "title": "Non-Consensual Synthetic Intimate Imagery: Prevalence, Attitudes, and Knowledge in 10 Countries",
      "publisher": "Umbach, Henry, Beard & Berryessa — Proceedings of CHI 2024 (arXiv preprint)",
      "source_type": "peer_reviewed",
      "statistic": "2.2% of all respondents indicated personal victimization with non-consensual synthetic intimate imagery; 1.8% indicated perpetration; survey of >16,000 respondents across 10 countries (Australia, Belgium, Denmark, France, Italy, Netherlands, Mexico, South Korea, UK, US)",
      "excerpt": "\"2.2% of all respondents indicated personal victimization, and 1.8% all of respondents indicated perpetration behaviors regarding deepfake pornography.\"\n",
      "source_date": "2024-02-02",
      "source_accessed": "2026-05-28",
      "archive_url": "http://web.archive.org/web/20260301234528/https://arxiv.org/abs/2402.01721",
      "calculation_notes": "Umbach et al. (2024) is the first peer-reviewed multi-country population survey specifically measuring synthetic (AI-generated) intimate-imagery victimization rather than the broader image-based-abuse category. The paper was accepted to CHI 2024 (the flagship ACM human-computer interaction conference). Sample is >16,000 respondents across the 10 listed countries; the 2.2% figure is the pooled personal-victimization rate across all respondents. This is the headline native rate. The fielding year is 2023, so the figure is best interpreted as \"ever experienced as of 2023\" — a contemporaneous lifetime-to-date prevalence with the technology only ~5 years into mainstream availability. The lifetime extrapolation to ~4% combines this base with a moderate forward-growth assumption (Home Security Heroes census shows 550% growth 2019-2023), capped well below the 8% lifetime rate Eaton et al. found for the broader NCII category that includes real imagery.\n"
    },
    {
      "url": "https://www.cybercivilrights.org/wp-content/uploads/2017/06/CCRI-2017-Research-Report.pdf",
      "title": "2017 Nationwide Online Study of Nonconsensual Porn Victimization and Perpetration",
      "publisher": "Eaton, Jacobs & Ruvalcaba — Cyber Civil Rights Initiative",
      "source_type": "primary_study",
      "statistic": "1 in 12 (8.0%) of US adult respondents reported lifetime victimization of nonconsensual pornography; 1 in 20 (5%) reported perpetration; women's victimization rate was higher than men's; n=3,044 US adults",
      "excerpt": "\"Among participants (54% women), 1 in 12 reported at least one instance of nonconsensual pornography victimization in their lifetime, and 1 in 20 reported perpetration of nonconsensual pornography. Women reported higher rates of victimization and lower rates of perpetration than men.\"\n",
      "source_date": "2017-06-12",
      "source_accessed": "2026-05-28",
      "archive_url": "http://web.archive.org/web/20260529100326/https://www.cybercivilrights.org/wp-content/uploads/2017/06/CCRI-2017-Research-Report.pdf",
      "calculation_notes": "The Eaton/CCRI 2017 study is the most-cited US adult NCII prevalence figure and was published as a peer-reviewed follow-up in Psychology of Violence (Ruvalcaba & Eaton, 2019). It covers all non-consensual intimate imagery — overwhelmingly authentic photographs taken consensually and later distributed without consent — not just AI-generated deepfakes, which were not yet a measurable category in 2016 when the survey was fielded. The 8% lifetime figure is included here as the broader-category ceiling: any plausible deepfake-specific lifetime rate should sit at or below this number, because synthetic imagery is one mechanism within a larger problem that pre-dates the AI tools. The CCRI sample (3,044 US adults via online panel) has the usual online-panel skew but is the most authoritative US-specific headline number available. Used to set the upper bound on the uncertainty interval, not the central estimate.\n",
      "independence_note": "Eaton et al. used a US-only online probability panel through CCRI, entirely separate from the Umbach et al. 10-country fielding. The broader definition (real + synthetic NCII) and the seven-year gap make the two figures complementary, not redundant: Umbach captures the synthetic share of a recent year; Eaton captures lifetime prevalence of the broader category before consumer deepfake tools existed.\n"
    },
    {
      "url": "https://www.siliconrepublic.com/enterprise/deepfakes-non-consensual-porn-research-deeptrace",
      "title": "96pc of deepfakes online are pornographic in nature (Deeptrace report coverage)",
      "publisher": "Silicon Republic — coverage of Deeptrace (Sensity) State of Deepfakes 2019",
      "source_type": "news_article",
      "statistic": "Deeptrace 2019 census found 14,678 deepfake videos online; 96% were non-consensual pornography; every observed pornographic deepfake portrayed a female subject",
      "excerpt": "\"Deepfake pornography is a phenomenon that exclusively targets and harms women.\"\n",
      "source_date": "2019-10-08",
      "source_accessed": "2026-05-28",
      "archive_url": "http://web.archive.org/web/20250115060625/https://www.siliconrepublic.com/enterprise/deepfakes-non-consensual-porn-research-deeptrace",
      "calculation_notes": "The Deeptrace (later renamed Sensity AI) 2019 report is the earliest systematic census of deepfake videos online and established the gender asymmetry that has held in every subsequent measurement: pornographic deepfakes overwhelmingly target women. Used here purely to source the gender multiplier — the headline native rate is from Umbach, the severity asymmetry is from Deeptrace. The original report PDF is published on Sensity's site; this Silicon Republic coverage is used because the primary host returns 403 to automated retrieval. The 96% pornographic / ~100% female-target finding has been replicated by Home Security Heroes (2023) on a much larger 95,820-video sample with 99% female-target.\n"
    },
    {
      "url": "https://www.securityhero.io/state-of-deepfakes/",
      "title": "2023 State of Deepfakes: Realities, Threats, and Impact",
      "publisher": "Home Security Heroes",
      "source_type": "reputable_reference",
      "statistic": "95,820 deepfake videos identified online in 2023 (a 550% increase from 2019); 98% of deepfake videos online are pornographic; 99% of deepfake pornography targets women",
      "excerpt": "\"Deepfake porn makes up 98 percent of all deepfake videos online, with 99 percent of them targeting women.\"\n",
      "source_date": "2023-09-01",
      "source_accessed": "2026-05-28",
      "archive_url": "http://web.archive.org/web/20260527182119/https://www.securityhero.io/state-of-deepfakes/",
      "calculation_notes": "The 2023 census is included for trajectory rather than base-rate: the 550% growth from 2019 to 2023 (14,678 → 95,820 detected videos) is the diffusion signal that drives the upward skew on the uncertainty band. Home Security Heroes is a commercial security publication rather than peer-reviewed academia, but its census methodology mirrors Deeptrace's 2019 work and the corroborating 99% female-target figure aligns with both the academic literature and the Sensity follow-ups. Used to justify treating the 2.2% contemporary prevalence as a lower bound on true lifetime exposure given the rapid capability and accessibility growth.\n"
    }
  ],
  "comparison_anchors": [
    {
      "label": "Lifetime NCII victimization (any kind, US adults, Eaton 2017)",
      "lifetime_us_adult": 0.08
    },
    {
      "label": "Image-based abuse victimization (Australia, eSafety 2017)",
      "lifetime_us_adult": 0.23
    },
    {
      "label": "Identity theft (US adult, lifetime)",
      "lifetime_us_adult": 0.3
    }
  ],
  "personal_factor_multipliers": [
    {
      "factor": "Woman (vs man, baseline)",
      "multiplier": 4,
      "notes": "Deeptrace 2019 found pornographic deepfakes were exclusively female-targeted; Home Security Heroes 2023 found 99% targeted women on a 95,820-video census; Umbach et al. 2024 reported higher victimization among women but did not publish the exact split in the abstract. The 4x multiplier averages the gender skew across all categories of synthetic intimate imagery (not just the most-replicated pornographic videos, where the female share approaches 100%)."
    },
    {
      "factor": "Public-facing online presence (creator, journalist, politician)",
      "multiplier": 3,
      "notes": "The Markup / DeepMedia 2024 analysis found 1 in 6 US Congresswomen had been targeted with sexually explicit AI imagery — roughly 16%, several multiples of the general-population base rate. Public-figure status materially elevates targeting because reference images are abundant and the social payoff to perpetrators is higher."
    },
    {
      "factor": "Young adult (18-29)",
      "multiplier": 2,
      "notes": "eSafety Commissioner (Australia, 2017) found 24% of women aged 18-24 had experienced image-based abuse vs 15% of all women; the age skew toward young adults in image-based abuse generally is documented across multiple national surveys (Henry, Powell & Flynn 2017; Eaton et al. 2017). Synthetic-imagery victimization is expected to follow the same age gradient because young adults' photographic footprint online is largest."
    },
    {
      "factor": "K-12 or higher-ed student in a community where peer-generated deepfakes have been reported",
      "multiplier": 5,
      "notes": "Internet Watch Foundation and multiple US/UK school incidents in 2023-2024 documented clusters of student-on-student synthetic intimate imagery — typically through 'nudify' apps applied to school yearbook or social-media photographs. The base rate inside such cohorts is materially elevated. Treated as a setting multiplier rather than a personal trait."
    }
  ],
  "short_label": "Intimate deepfake",
  "myth_framing": "underrated",
  "outcome_severity": "serious_harm",
  "exposure_pattern": "acute",
  "outcome_type": "mental_trauma",
  "valence": "negative",
  "caveats": "Three caveats matter more than the headline number. First, the gender asymmetry is severe enough that the population-average figure misleads: the 4% lifetime estimate is roughly a weighted average of a female rate near 7-8% and a male rate well under 2%. Every census of deepfake-video supply finds the pornographic share targets women essentially exclusively. Reporting only the population average would understate the risk for women and overstate it for men, which is why the personal-factor multipliers are load-bearing rather than decorative.\nSecond, the survey data is almost certainly conservative. Many victims of synthetic intimate imagery never learn the imagery exists — it can be generated, shared, and consumed entirely outside the subject's social graph. The 2.2% contemporary figure measures self-aware victimization; the true rate is necessarily higher and unknowable by self-report. This is structurally different from cyberbullying or harassment, where the victim is by definition aware of the event.\nThird, no comparable peer-reviewed time series exists. Umbach et al. (2024) is the first multi-country population survey on this specific category. Deeptrace 2019, Home Security Heroes 2023, and the Internet Watch Foundation reports measure supply (videos online) rather than demand-side victim prevalence. The uncertainty band reflects both measurement gaps and the steepness of the technology adoption curve — any number computed from 2023 data may be substantially low by 2030.\n",
  "quality_score": {
    "d1": 4,
    "d2": 5,
    "d3": 3,
    "d4": 4,
    "d5": 4,
    "d6": 4,
    "d7": 4,
    "d8": 5,
    "avg": 4.125,
    "scored_by": "claude-code-8d",
    "scored_at": "2026-05-28",
    "methodology_version": "1.2"
  },
  "reviewer": "8d-eval-2026-05-28",
  "last_reviewed": "2026-05-28",
  "reviewed": true,
  "generated_at": "2026-05-28",
  "image": {
    "alt": "A single closed laptop on a plain desk surface with a small blue notification dot visible at the lid edge, flat vector illustration in muted tones."
  },
  "attribution": "Likelier — https://likelier.app",
  "license": "https://creativecommons.org/licenses/by-sa/4.0/",
  "support": "https://buymeacoffee.com/kgluszczyk?via=likelier&utm_content=api-fear-single",
  "canonical_url": "https://likelier.app/ai-deepfake-intimate-image-adult"
}