---
type: source
id: miller-2023-ai-hyperrealism
title: 'AI hyperrealism: Why AI faces are perceived as more real than human ones'
authors:
- Elizabeth J. Miller
- Ben A. Steward
- Zak Witkower
- Clare A. M. Sutherland
- Eva G. Krumhuber
- Amy Dawel
year: 2023
venue: Psychological Science
volume: 34(12)
pages: 1390-1403
doi: 10.1177/09567976231207095
peer_review: peer-reviewed-journal
evidence_tier: B
design: Reanalysis of Nightingale & Farid (2022) data; Exp1 real-vs-AI judgments with confidence (White faces, White US participants); Exp2 ratings of 14 face attributes + lens-model analysis and random-forest classifier
modality:
- image-face
stimuli: White and non-White StyleGAN2 faces and human faces from the Nightingale & Farid (2022) set
n_participants: 734
n_note: Exp1 n=124 (White US adults); Exp2 n=610; reanalysis ≈315 (Nightingale & Farid Exp1)
topics:
- human-detection
- hyperrealism
- overconfidence
- uncanny-valley
summary: White AI faces were judged human 65.9% of the time vs 51.1% for real White faces ('AI hyperrealism'); the worst detectors were the most confident; a classifier on human-rated attributes reached 94%.
results:
- metric: 'Exp1: White AI faces judged human'
  value: 65.9%
  note: ''
  source_location: null
- metric: 'Exp1: White human faces judged human'
  value: 51.1%
  note: ''
  source_location: null
- metric: 'Reanalysis of Nightingale & Farid: White AI vs human judged human'
  value: 69.5% vs 52.2%
  note: non-White faces ≈ chance, no significant AI-human difference
  source_location: null
- metric: Confidence vs errors (AI faces)
  value: r = .385
  note: more errors ↔ higher confidence (Dunning-Kruger-like)
  source_location: null
- metric: Negative meta-d'
  value: 59% of participants
  note: poor insight into own errors
  source_location: null
- metric: Random forest on 14 human-rated attributes
  value: 94% [95% CI 91, 97]
  note: cross-validated; cues exist but humans misuse them
  source_location: null
moderators:
- Participants misread facial proportions, familiarity and memorability as signs of humanness; attractiveness and symmetry were used correctly.
- Authors attribute White-only hyperrealism to StyleGAN2 training data being ≈69% White, making White AI faces more average.
limitations:
- White participants only in Exp1
- StyleGAN2 only
- Stimuli reused from Nightingale & Farid
verification:
  status: verified-fulltext
  confidence: high
  checked: 2026-10-07
  via: journals.sagepub.com article page
url: null
pmid: null
pmcid: null
arxiv_id: null
openalex_id: null
oa_url: null
cited_in:
- MR-0001
- MR-0002
updated: '2026-10-07'
---

# AI hyperrealism: Why AI faces are perceived as more real than human ones

> **Source record** `miller-2023-ai-hyperrealism` · 2023 · Psychological Science  
> **Evidence tier:** B (peer-reviewed-journal) · **Verification:** verified-fulltext  
> **Bottom line:** White AI faces were judged human 65.9% of the time vs 51.1% for real White faces ('AI hyperrealism'); the worst detectors were the most confident; a classifier on human-rated attributes reached 94%.

## Citation {#citation}

Elizabeth J. Miller, Ben A. Steward, Zak Witkower, Clare A. M. Sutherland, Eva G. Krumhuber, Amy Dawel (2023). AI hyperrealism: Why AI faces are perceived as more real than human ones. *Psychological Science*, 34(12), 1390-1403. https://doi.org/10.1177/09567976231207095

Verification entry (identifiers, resolvers, claims with source locations): [/data/citations.json#miller-2023-ai-hyperrealism](/data/citations.json#miller-2023-ai-hyperrealism)

## Study design {#design}

| Field | Value |
|---|---|
| Design | Reanalysis of Nightingale & Farid (2022) data; Exp1 real-vs-AI judgments with confidence (White faces, White US participants); Exp2 ratings of 14 face attributes + lens-model analysis and random-forest classifier |
| Modality | image-face |
| Stimuli | White and non-White StyleGAN2 faces and human faces from the Nightingale & Farid (2022) set |
| Participants (N) | 734 |
| N detail | Exp1 n=124 (White US adults); Exp2 n=610; reanalysis ≈315 (Nightingale & Farid Exp1) |
| Topics | human-detection, hyperrealism, overconfidence, uncanny-valley |

## Results {#results}

| Metric | Value | Note |
|---|---|---|
| Exp1: White AI faces judged human | 65.9% |  |
| Exp1: White human faces judged human | 51.1% |  |
| Reanalysis of Nightingale & Farid: White AI vs human judged human | 69.5% vs 52.2% | non-White faces ≈ chance, no significant AI-human difference |
| Confidence vs errors (AI faces) | r = .385 | more errors ↔ higher confidence (Dunning-Kruger-like) |
| Negative meta-d' | 59% of participants | poor insight into own errors |
| Random forest on 14 human-rated attributes | 94% [95% CI 91, 97] | cross-validated; cues exist but humans misuse them |

## Moderators {#moderators}

- Participants misread facial proportions, familiarity and memorability as signs of humanness; attractiveness and symmetry were used correctly.
- Authors attribute White-only hyperrealism to StyleGAN2 training data being ≈69% White, making White AI faces more average.

## Limitations {#limitations}

- White participants only in Exp1
- StyleGAN2 only
- Stimuli reused from Nightingale & Farid

## Cited in {#cited-in}

- [MR-0001](/metareports/human-detection-accuracy.md): Human Detection of Deepfakes: Accuracy, Bias, Confidence, and Training
- [MR-0002](/metareports/uncanny-valley-and-synthetic-faces.md): The Uncanny Valley and Synthetic Media: Does It Exist, What Causes It, and Have Deepfakes Crossed It?

## Verification {#verification}

| Field | Value |
|---|---|
| Status | `verified-fulltext` |
| Compiler confidence | high |
| Checked on | 2026-10-07 |
| Via | journals.sagepub.com article page |