---
type: source
id: sulimani-2026-recognition-ai-faces
title: Recognition of real and artificial intelligence-generated faces
authors:
- N. Sulimani
- A. van Hunenstijn
- A. Albonico
year: 2026
venue: Perception
volume: 55(10)
pages: 924-937
doi: 10.1177/03010066261442089
pmid: '42053250'
peer_review: peer-reviewed-journal
evidence_tier: B
design: Face-memory (recognition) task with real and AI-generated faces, plus a human-vs-AI classification task; correlation between recognition and classification accuracy
modality:
- image-face
stimuli: Real and AI-generated faces (generator not in abstract)
n_participants: null
n_note: not in abstract
topics:
- face-recognition
- memory
- human-detection
- covert-recognition
summary: AI faces were recognised (remembered) more accurately than real faces, both face types were classified above chance with no difference, and recognition and classification accuracy were not correlated.
results:
- metric: Recognition memory
  value: significantly higher for AI faces than real faces
  note: no values in abstract
  source_location: abstract
- metric: Classification
  value: above chance for both face types; no difference between them
  note: ''
  source_location: abstract
- metric: Recognition vs classification
  value: not significantly correlated
  note: authors suggest partial unawareness of the AI-face memory advantage
  source_location: abstract
moderators: []
limitations:
- Values not checked against full text
- Generator and N not recorded
verification:
  status: verified-abstract
  confidence: medium
  checked: 2026-10-18
  via: Europe PMC abstract record (PMID 42053250)
url: null
pmcid: null
arxiv_id: null
openalex_id: null
oa_url: null
cited_in:
- MR-0001
- MR-0002
updated: '2026-10-18'
---

# Recognition of real and artificial intelligence-generated faces

> **Source record** `sulimani-2026-recognition-ai-faces` · 2026 · Perception  
> **Evidence tier:** B (peer-reviewed-journal) · **Verification:** verified-abstract (compiler confidence: medium)  
> **Bottom line:** AI faces were recognised (remembered) more accurately than real faces, both face types were classified above chance with no difference, and recognition and classification accuracy were not correlated.

## Citation {#citation}

N. Sulimani, A. van Hunenstijn, A. Albonico (2026). Recognition of real and artificial intelligence-generated faces. *Perception*, 55(10), 924-937. https://doi.org/10.1177/03010066261442089

Verification entry (identifiers, resolvers, claims with source locations): [/data/citations.json#sulimani-2026-recognition-ai-faces](/data/citations.json#sulimani-2026-recognition-ai-faces)

## Study design {#design}

| Field | Value |
|---|---|
| Design | Face-memory (recognition) task with real and AI-generated faces, plus a human-vs-AI classification task; correlation between recognition and classification accuracy |
| Modality | image-face |
| Stimuli | Real and AI-generated faces (generator not in abstract) |
| Participants (N) | not recorded |
| N detail | not in abstract |
| Topics | face-recognition, memory, human-detection, covert-recognition |

## Results {#results}

| Metric | Value | Note |
|---|---|---|
| Recognition memory | significantly higher for AI faces than real faces | no values in abstract |
| Classification | above chance for both face types; no difference between them |  |
| Recognition vs classification | not significantly correlated | authors suggest partial unawareness of the AI-face memory advantage |

## Limitations {#limitations}

- Values not checked against full text
- Generator and N not recorded

## 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-abstract` |
| Compiler confidence | medium |
| Checked on | 2026-10-18 |
| Via | Europe PMC abstract record (PMID 42053250) |

Values on this page have not all been checked against the full text. Confirm them against the primary source before high-stakes use.