Synthetic ImprintingReference Project
Exposure

Voice Cloning

Voice cloning uses recorded speech to generate new audio resembling a particular person's voice. It may synthesize new speech or transform an existing recording toward a target speaker's characteristics. The resemblance is a technical output, not proof that the person spoke or approved the words. A generic synthetic voice is not necessarily an individual clone.

Term statusEmerging Concept

A term in research or professional use whose meaning still varies across fields. About these labels.

Evidence for the claims on this page: Technical demonstration

The cited system demonstrates voice synthesis from reference recordings. It does not measure listener trust, memory, or the effects of misuse.

How to read this assessment

By SyntheticImprinting.com Editorial ProjectUpdated Revision history

On this page

What it is

Voice cloning is the synthesis or conversion of speech to resemble an identifiable target speaker. Its quality depends on the system, reference recordings, language, and conditions in which the output is heard. (Jia et al., 2018)

In everyday language

A system learns features of a voice from examples and uses them to produce speech the person may never have recorded.

What it isn't

A generic synthetic voice is not necessarily a clone of an individual. A convincing resemblance does not prove that the person spoke, approved, or even knew about the words.

An example

Hypothetical example: with a narrator's permission, a publisher uses a modeled version of that narrator's voice to produce an updated introduction.

Key features

  • Targets characteristics of a particular voice
  • Can produce new spoken material
  • Requires a distinction between vocal resemblance and a genuine recording

How it relates to synthetic imprinting

Recognizable speech is one form of synthetic exposure. Researchers can investigate whether voice familiarity changes credibility or later source judgments.

Research context

Jia and colleagues demonstrated a system that synthesizes speech in voices learned from reference audio. A technical demonstration of resemblance does not establish psychological effects or consent to a particular use. (Jia et al., 2018)

Emerging Concept

Deepfake

A deepfake is realistic synthetic or manipulated audio, imagery, or video that depicts a person or event differently from what actually occurred.

Emerging Concept

Synthetic Media

Synthetic media is content created or substantially changed by software, including AI-generated images, voices, video, and text.

Established Concept

Source Monitoring

Source monitoring is the process of judging where a memory or piece of information came from, such as a conversation, an image, or direct experience.

Sources and what they support

  1. Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis — Ye Jia, Yu Zhang, Ron J. Weiss, Quan Wang, Jonathan Shen, Fei Ren, Zhifeng Chen, Patrick Nguyen, Ruoming Pang, Ignacio Lopez Moreno, Yonghui Wu (2018). Advances in Neural Information Processing Systems 31.

    Peer-reviewed research. Describes a speech-synthesis system that adapts to speaker characteristics from reference recordings; this is a technical study, not a study of listener trust or memory.

Cite this definition

SyntheticImprinting.com Editorial Project. (2026-09-27). Voice Cloning. Version 1.4. SyntheticImprinting.com. https://syntheticimprinting.com/terms/voice-cloning/

The version identifies the page revision. This address opens the current page; it is not an immutable archive. You can also select and copy the reference above.

APA: SyntheticImprinting.com. (2026). Voice Cloning. https://syntheticimprinting.com/terms/voice-cloning/

MLA: “Voice Cloning.” SyntheticImprinting.com, 2026, https://syntheticimprinting.com/terms/voice-cloning/. Accessed November 12, 2026.

Chicago: SyntheticImprinting.com. “Voice Cloning.” Last modified 2026-09-27. https://syntheticimprinting.com/terms/voice-cloning/.

Revision history

These are retained versions of the opening definition. Revision dates describe editorial work, not independent scientific validation.

  1. Version 1.4 · Current

    Added claim-level citations and a separate assessment of evidence support. The opening definition is unchanged.

    Read this version’s definition
    Voice cloning uses recorded speech to generate new audio resembling a particular person's voice. It may synthesize new speech or transform an existing recording toward a target speaker's characteristics. The resemblance is a technical output, not proof that the person spoke or approved the words. A generic synthetic voice is not necessarily an individual clone.
  2. Version 1.3

    Expanded the opening definition; added explicit boundaries, source context, related-term cards, and linked editorial attribution.

    Read the previous definition
    Voice cloning uses recorded speech to generate new audio resembling a particular person's voice. It may synthesize new speech or transform an existing recording toward a target speaker's characteristics. The resemblance is a technical output, not proof that the person spoke or approved the words. A generic synthetic voice is not necessarily an individual clone.
  3. Version 1.0

    Definition recorded in the previous release. Earlier edits are not reconstructed here.

    Read the previous definition
    Voice cloning uses recorded speech to generate new audio resembling a particular person's voice.

What the status means

Emerging Concept

A term in research or professional use whose meaning still varies across fields.

How evidence labels work
First published
Updated
Version
1.4
Review status
External expert review pending

Help improve this entry

Suggest a correction to Voice Cloning. Your email app opens with the page address and version included. Add the passage, your suggested change, and a source if you have one.

You can also email contact@syntheticimprinting.com directly.