---
type: source
id: nightingale-farid-2022
title: AI-synthesized faces are indistinguishable from real faces and more trustworthy
authors:
- Sophie J. Nightingale
- Hany Farid
year: 2022
venue: Proceedings of the National Academy of Sciences (PNAS)
volume: 119(8)
pages: e2120481119
doi: 10.1073/pnas.2120481119
pmcid: PMC8872790
peer_review: peer-reviewed-journal
evidence_tier: B
design: '3 online experiments: real-vs-synthetic classification (Exp1), same with training + trial-by-trial feedback (Exp2), trustworthiness ratings 1-7 (Exp3)'
modality:
- image-face
stimuli: 400 StyleGAN2 faces + 400 real faces matched 1:1 from a 70,000-face database (FFHQ-style); balanced 200 women/200 men; 100 each Black, White, East Asian, South Asian
n_participants: 757
n_note: Exp1 n=315; Exp2 n=219; Exp3 n=223; each judged 128 faces
topics:
- human-detection
- hyperrealism
- trust
- training
summary: StyleGAN2 faces were classified at chance (48.2%); training+feedback raised accuracy only to 59.0%; synthetic faces were rated 7.7% more trustworthy than real faces.
results:
- metric: Exp1 accuracy (no training)
  value: 48.2% [95% CI 47.1, 49.2]
  note: d'=-0.09, beta=0.99 (no response bias)
  source_location: null
- metric: Exp2 accuracy (training + feedback)
  value: 59.0% [95% CI 57.7, 60.4]
  note: d'=0.46; no improvement across session (first half 59.3%, second half 58.8%)
  source_location: null
- metric: Exp3 trustworthiness (1-7)
  value: synthetic 4.82 vs real 4.48
  note: +7.7% for synthetic; t(222)=14.6, p<0.001, d=0.49
  source_location: null
- metric: Gender effect on trust
  value: women 4.94 vs men 4.36
  note: d=0.82
  source_location: null
- metric: Smiling confound check
  value: 65.5% real vs 58.8% synthetic faces smiling
  note: smiling does not explain higher synthetic trust
  source_location: null
moderators:
- White synthetic faces were classified least accurately; male White faces worse than female White faces. Authors attribute this to over-representation of White faces in StyleGAN2 training data.
- Feedback produced a one-off gain but no within-session learning, implying some synthetic faces carry no detectable artifacts.
limitations:
- Single generator (StyleGAN2)
- Curated, artifact-screened stimuli
- Context-free still faces
- Online samples
verification:
  status: verified-fulltext
  confidence: high
  checked: 2026-10-07
  via: Europe PMC full text PMC8872790
url: null
pmid: null
arxiv_id: null
openalex_id: null
oa_url: null
cited_in:
- MR-0001
- MR-0002
updated: '2026-10-07'
---

# AI-synthesized faces are indistinguishable from real faces and more trustworthy

> **Source record** `nightingale-farid-2022` · 2022 · Proceedings of the National Academy of Sciences (PNAS)  
> **Evidence tier:** B (peer-reviewed-journal) · **Verification:** verified-fulltext  
> **Bottom line:** StyleGAN2 faces were classified at chance (48.2%); training+feedback raised accuracy only to 59.0%; synthetic faces were rated 7.7% more trustworthy than real faces.

## Citation {#citation}

Sophie J. Nightingale, Hany Farid (2022). AI-synthesized faces are indistinguishable from real faces and more trustworthy. *Proceedings of the National Academy of Sciences (PNAS)*, 119(8), e2120481119. https://doi.org/10.1073/pnas.2120481119

Verification entry (identifiers, resolvers, claims with source locations): [/data/citations.json#nightingale-farid-2022](/data/citations.json#nightingale-farid-2022)

## Study design {#design}

| Field | Value |
|---|---|
| Design | 3 online experiments: real-vs-synthetic classification (Exp1), same with training + trial-by-trial feedback (Exp2), trustworthiness ratings 1-7 (Exp3) |
| Modality | image-face |
| Stimuli | 400 StyleGAN2 faces + 400 real faces matched 1:1 from a 70,000-face database (FFHQ-style); balanced 200 women/200 men; 100 each Black, White, East Asian, South Asian |
| Participants (N) | 757 |
| N detail | Exp1 n=315; Exp2 n=219; Exp3 n=223; each judged 128 faces |
| Topics | human-detection, hyperrealism, trust, training |

## Results {#results}

| Metric | Value | Note |
|---|---|---|
| Exp1 accuracy (no training) | 48.2% [95% CI 47.1, 49.2] | d'=-0.09, beta=0.99 (no response bias) |
| Exp2 accuracy (training + feedback) | 59.0% [95% CI 57.7, 60.4] | d'=0.46; no improvement across session (first half 59.3%, second half 58.8%) |
| Exp3 trustworthiness (1-7) | synthetic 4.82 vs real 4.48 | +7.7% for synthetic; t(222)=14.6, p<0.001, d=0.49 |
| Gender effect on trust | women 4.94 vs men 4.36 | d=0.82 |
| Smiling confound check | 65.5% real vs 58.8% synthetic faces smiling | smiling does not explain higher synthetic trust |

## Moderators {#moderators}

- White synthetic faces were classified least accurately; male White faces worse than female White faces. Authors attribute this to over-representation of White faces in StyleGAN2 training data.
- Feedback produced a one-off gain but no within-session learning, implying some synthetic faces carry no detectable artifacts.

## Limitations {#limitations}

- Single generator (StyleGAN2)
- Curated, artifact-screened stimuli
- Context-free still faces
- Online samples

## 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 | Europe PMC full text PMC8872790 |