{
  "slug": "ai-voice-clone-scam-target",
  "question": "What are the odds you'll be targeted by an AI voice-clone scam in your lifetime?",
  "category": "tech",
  "tags": [
    "digital-fraud",
    "elder-care"
  ],
  "no_reliable_estimate": false,
  "perceived": {
    "description": "Public coverage of AI voice cloning has been dominated by single dramatic incidents — the Arizona mother who heard her daughter's cloned voice begging for ransom in 2023, the $25 million Hong Kong deepfake video call in 2024, the impostor impersonating Secretary of State Marco Rubio in 2025 — which leaves the typical reader with two competing intuitions: that this is everywhere and that it is exotic. When asked to estimate how many adults have already received a voice-clone scam attempt directed at them personally, guesses cluster around 1–3% — roughly the rate of dramatic news stories per capita. The measured figure from McAfee's 2023 seven-country survey is about 10% of adults already personally targeted at the time of the survey, with another 15% saying it had happened to someone they know. The actual ceiling, integrated over a full adult lifetime in 2026 onward, is meaningfully higher than the snapshot suggests.\n",
    "rough_estimate": "~1–3% lifetime feels right to most respondents",
    "kind": "intuition"
  },
  "native": {
    "display": "~10 in 100 adults personally targeted by an AI voice-clone scam (point prevalence, McAfee 2023, 7 countries)",
    "numerator": 10,
    "denominator": 100,
    "unit": "personally targeted to date",
    "population": "adults in McAfee's 7-country sample (US, UK, France, Germany, Japan, Australia, India)"
  },
  "normalized": {
    "lifetime_us_adult": 0.35,
    "display": "~35% lifetime probability of being targeted by an AI voice-clone scam (US adult)",
    "log_value": -0.46,
    "assumptions": "McAfee's April 2023 survey (N=7,054 across 7 countries, MSI Research panel) found that 10% of respondents had personally been targeted by a voice-clone scam, with another 15% saying someone they knew had been. That 10% is a point-prevalence snapshot covering the first ~18 months of widely available consumer voice-cloning tools (ElevenLabs launched in beta January 2023; open-source models followed within months). The lifetime figure extrapolates from that snapshot under three forces: (1) the underlying tool capability is maturing rapidly — the FBI's December 2024 IC3 PSA confirmed that voice cloning is now a standard tool in scam campaigns alongside text and image generation; (2) impersonation scams are already the FTC's third-largest complaint category, with Americans losing close to $3 billion to imposter scams in 2024; (3) the per-year targeting hazard is bounded below by the McAfee 10% / ~1.5 years ≈ 6–7% per year and above by FTC imposter-scam complaint rates (~1% per year reporting an actual loss, with many more contacted but not defrauded). A central estimate of 35% lifetime sits between the naive compound of a 1% annual targeting hazard over 59 years (~45%) and the floor from McAfee's already-observed 10% prevalence in 2023. The uncertainty band (0.18–0.55) reflects (a) wide variation in what counts as \"targeted\" (a fully personalised cloned-voice call versus a generic AI-voiced robocall), (b) demographic targeting concentrated among older adults and people with public audio footprints (podcasters, executives), and (c) the rapid technology trajectory making point estimates inherently fragile beyond a 5-year horizon.\n",
    "uncertainty": {
      "low": 0.18,
      "high": 0.55
    },
    "scope": "us_adult_lifetime"
  },
  "sources": [
    {
      "url": "https://www.mcafee.com/ai/news/ai-voice-scam/",
      "title": "Scammers Use AI Voice Cloning Tools to Fuel New Scams",
      "publisher": "McAfee",
      "source_type": "reputable_reference",
      "statistic": "25% of adults have personally experienced or know someone who has experienced an AI voice scam (10% personally, 15% know someone); 77% of those targeted lost money; 70% of adults not confident they could distinguish a cloned voice from the real one",
      "excerpt": "\"A quarter of adults surveyed have previously experienced some kind of AI voice scam, with 1 in 10 targeted personally and 15% saying it happened to someone they know. 77% of victims said they had lost money as a result. [...] 70% of adults are not confident that they could identify the cloned version from the real thing.\"\n",
      "source_date": "2023-05-02",
      "source_accessed": "2026-05-28",
      "archive_url": "http://web.archive.org/web/20260531014325/https://www.mcafee.com/ai/news/ai-voice-scam/",
      "calculation_notes": "The 10% personally-targeted figure is the spine of the native estimate (10 in 100). It is a global average across 7 countries (US, UK, France, Germany, Japan, Australia, India). The US-only subsample is not separately reported by McAfee; US adults are assumed to face at least the global rate given that English-language voice cloning is the most mature subset of the technology and that US adults receive the highest volume of impersonation-scam contact attempts per FTC Sentinel data. The 77% loss rate among targeted individuals is conditional on responding to the scam, not on receiving the call; the unconditional \"targeted AND lost money\" rate is therefore lower than 77% of 10%, since not every targeted adult engages with the scammer.\n",
      "independence_note": "Commissioned online survey conducted by MSI Research between April 13 and April 19, 2023. Methodologically independent from the FTC and FBI complaint databases.\n"
    },
    {
      "url": "https://www.ic3.gov/PSA/2024/PSA241203",
      "title": "Criminals Use Generative Artificial Intelligence to Facilitate Financial Fraud",
      "publisher": "FBI Internet Crime Complaint Center (IC3)",
      "source_type": "govt_report",
      "statistic": "FBI public service announcement identifying voice cloning as one of five primary generative-AI fraud vectors in active use; no aggregate incident count provided",
      "excerpt": "\"Generative AI reduces the time and effort criminals must expend to deceive their targets. [...] Criminals generate short audio clips containing a loved one's voice to impersonate a close relative in a crisis situation, asking for immediate financial assistance or demanding a ransom.\"\n",
      "source_date": "2024-12-03",
      "source_accessed": "2026-05-28",
      "archive_url": "http://web.archive.org/web/20260515181110/https://www.ic3.gov/PSA/2024/PSA241203",
      "calculation_notes": "The FBI PSA confirms that voice cloning has moved from research curiosity to standard criminal tool by late 2024, supporting the assumption that the per-year targeting hazard is non-trivial and rising rather than rare and stable. The PSA does not publish counts of voice-clone-specific complaints because IC3 categorises by scam type (grandparent scam, tech-support scam, romance scam), not by impersonation technique. This means voice-clone incidents are scattered across multiple categories and almost certainly undercounted in any single line of IC3 reporting.\n",
      "independence_note": "Federal law enforcement advisory. Independent from McAfee's commercial survey and from FTC consumer-complaint data.\n"
    },
    {
      "url": "https://advocacy.consumerreports.org/press_release/more-than-75000-consumers-urge-ftc-to-crack-down-on-ai-voice-cloning-fraud/",
      "title": "More Than 75,000 Consumers Urge FTC to Crack Down on AI Voice Cloning Fraud",
      "publisher": "Consumer Reports",
      "source_type": "reputable_reference",
      "statistic": "Americans lost nearly $3 billion in imposter scams during 2024; six commercially available AI voice-cloning tools tested lacked meaningful safeguards against misuse",
      "excerpt": "\"Americans lost nearly $3 billion in imposter scams during 2024 alone. [...] AI voice cloning tools are making it easier than ever for scammers to impersonate someone's voice. [...] Consumer Reports assessed six freely or cheaply available AI voice cloning products and found that most did not have meaningful safeguards to stop fraud or misuse.\"\n",
      "source_date": "2025-08-13",
      "source_accessed": "2026-05-28",
      "archive_url": "http://web.archive.org/web/20260210220448/https://advocacy.consumerreports.org/press_release/more-than-75000-consumers-urge-ftc-to-crack-down-on-ai-voice-cloning-fraud",
      "calculation_notes": "The $3 billion / 2024 imposter-scam total is the FTC Consumer Sentinel envelope inside which voice-clone scams sit as one technique. Voice-clone-specific losses are not separately published, but the magnitude and growth of the parent category establish that the targeting attempt rate must be much higher than the loss rate (most contacted adults do not engage or do not pay). This source also documents the supply side — that the underlying tools remain cheap and accessible — which is the mechanism behind the rising-hazard assumption in the normalized calculation.\n",
      "independence_note": "Consumer Reports advocacy filing to the FTC, citing FTC Sentinel data and Consumer Reports' own product testing. Independent from McAfee survey and FBI advisory.\n"
    },
    {
      "url": "https://www.aarp.org/pri/topics/work-finances-retirement/fraud-consumer-protection/ai-fraud-concerns-older-adults/",
      "title": "Older Adults Express High Concern and Limited Knowledge About AI Scams and Fraud",
      "publisher": "AARP Research",
      "source_type": "reputable_reference",
      "statistic": "84% of US adults 50+ are concerned about criminals using AI for voice cloning; 77% worry about becoming personal targets of AI-related fraud (N=1,000, Foresight 50+ Omnibus panel, fielded August 15–19, 2024)",
      "excerpt": "\"Older adults are worried about criminals using AI for a variety of fraudulent activities. [...] Voice cloning: 84 percent [are concerned]. [...] A significant majority of older adults expressed worry that they might personally become targets of an AI-related fraud in the future (77 percent).\"\n",
      "source_date": "2024-10-01",
      "source_accessed": "2026-05-28",
      "archive_url": "http://web.archive.org/web/20260120195848/https://www.aarp.org/pri/topics/work-finances-retirement/fraud-consumer-protection/ai-fraud-concerns-older-adults/",
      "calculation_notes": "Used to characterise the older-adult subgroup that is disproportionately targeted and contributes to the personal_factor_multipliers. The 84% concern figure tracks worry, not incidence, and is used only on the perception side and as evidence that voice-clone scams have moved into mainstream awareness. AARP did not publish an incidence figure for personal voice-clone targeting in this wave, so the calculation remains anchored on McAfee.\n",
      "independence_note": "Nationally representative US online/phone panel of adults 50+. Methodologically independent from McAfee global commercial survey, FBI complaint database, and FTC Sentinel.\n"
    }
  ],
  "comparison_anchors": [
    {
      "label": "Online scam loss (lifetime, US adult)",
      "lifetime_us_adult": 0.4
    },
    {
      "label": "Identity theft (lifetime, US adult)",
      "lifetime_us_adult": 0.6
    },
    {
      "label": "Investment scam loss — pig-butchering era (lifetime, US adult)",
      "lifetime_us_adult": 0.02
    }
  ],
  "personal_factor_multipliers": [
    {
      "factor": "Age 65+",
      "multiplier": 1.5,
      "notes": "Older adults are the most frequently impersonated relatives in grandparent-style voice-clone scams and the demographic with the highest median loss per imposter incident per FTC Sentinel 2024"
    },
    {
      "factor": "Public audio footprint (podcaster, executive, public speaker)",
      "multiplier": 3,
      "notes": "Voice cloning currently requires only 3 seconds of clean audio per McAfee 2023 and ElevenLabs documentation; people with hours of public recordings are first-target candidates and were the basis of the 2024 Hong Kong deepfake CEO case"
    },
    {
      "factor": "Family member with public audio footprint",
      "multiplier": 2,
      "notes": "Targeting is often indirect — the scammer clones a relative's voice to call the wealthier or more trusting family member. AARP 2024 documents elder relatives as the most common payment-side targets even when the cloned voice belongs to a younger family member"
    },
    {
      "factor": "No landline / unlisted mobile / no social media",
      "multiplier": 0.4,
      "notes": "Voice-clone scam delivery depends on the scammer obtaining both a contact number and a voice sample; adults with restricted social media presence and unpublished phone numbers have substantially lower targeting probability per general scam-exposure research"
    },
    {
      "factor": "Family safe-word established",
      "multiplier": 0.7,
      "notes": "A pre-arranged verbal challenge — recommended by the FBI's December 2024 PSA — does not reduce the targeting rate but reduces the conditional probability of a successful scam given targeting; included here because the headline framing is targeting probability, and a safe word changes downstream outcomes more than upstream contact"
    }
  ],
  "short_label": "AI voice scam",
  "myth_framing": "underrated",
  "outcome_severity": "moderate_harm",
  "exposure_pattern": "recurring",
  "outcome_type": "financial",
  "valence": "negative",
  "caveats": "\"Targeted\" is a deliberately broad construct: it covers receiving a voice-clone scam call or voicemail directed personally, whether or not the recipient engaged or lost money. The probability of being defrauded by such a scam is much lower than the probability of being targeted by one — McAfee's data implies roughly 77% of those who engaged with the scammer lost money, but most targeted adults hang up, ignore the voicemail, or call the supposed relative back through known channels. The single biggest definitional uncertainty is whether to count generic AI-voiced robocalls (which now make up a meaningful share of unsolicited call traffic) as \"voice-clone scams\" — this entry treats only personalised impersonation as a clone-scam attempt; the broader robocall category would push the lifetime figure substantially higher. The McAfee survey is the only large-sample incidence measure available and is now three years old in a fast-moving technology area; the FBI and FTC have confirmed that voice cloning is now a standard tool but have not published updated incidence counts. The 35% central estimate is therefore a model-based projection from a single 2023 snapshot and should be revised when comparable post-2025 surveys appear. Population heterogeneity is large: people with significant public audio footprints are at much higher risk than the population mean, and older adults face the highest median financial harm given a successful scam.\n",
  "quality_score": {
    "d1": 4,
    "d2": 4,
    "d3": 3,
    "d4": 3,
    "d5": 4,
    "d6": 4,
    "d7": 4,
    "d8": 4,
    "avg": 3.75,
    "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 smartphone resting face-up on a wooden kitchen table, screen showing only a blank incoming call, no caller ID, no people in the frame."
  },
  "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-voice-clone-scam-target"
}