Within Synthetic Media
When Real Images Become Easier to Deny
The liar's dividend shows how deepfakes can damage trust even when the disputed photo or video is genuine.
On this page
- What the liar's dividend means
- How synthetic media weakens genuine evidence
- What makes a denial credible or weak
Page outline Jump by section
Introduction
Artificial intelligence does not only create convincing fake images and videos. It also changes how people judge authentic ones. This secondary effect is often called the liar’s dividend: once the public knows that realistic synthetic media exists, a person confronted with genuine visual evidence can claim that it was generated or altered by AI. The result is a paradox. Improvements in fake media can weaken trust in real media. Researchers Robert Chesney and Danielle Citron, who popularised the term, argued that deepfakes may help people evade accountability for things that are actually true. [California Law Review]californialawreview.orgCalifornia Law ReviewDeep Fakes: A Looming Challenge for Privacy, Democracy…And, in what we call the “liar's dividend,” deep fakes mak…
Within the broader information-integrity challenge posed by synthetic images and videos, the liar’s dividend is important because it shifts attention away from deception alone. The problem is not merely that fake photographs can be believed. It is also that genuine photographs can be dismissed.
What the liar’s dividend means
The liar’s dividend describes the advantage gained by someone who falsely claims that authentic evidence is fabricated. The concept emerged from concerns that increasingly realistic AI-generated media would create uncertainty about what can be trusted online. [California Law Review]californialawreview.orgCalifornia Law ReviewDeep Fakes: A Looming Challenge for Privacy, Democracy…And, in what we call the “liar's dividend,” deep fakes mak…
The mechanism is straightforward:
- People become aware that convincing AI-generated images and videos exist.
- Trust in visual evidence declines.
- When damaging evidence appears, the subject can claim it is fake.
- Some observers become uncertain, even if the evidence is genuine.
The dividend is not the fake itself. The dividend is the benefit obtained from the possibility that a fake could exist. As awareness of deepfakes grows, the denial becomes more plausible to some audiences. Chesney and Citron noted that the phenomenon can grow alongside public education about deepfakes because greater awareness of manipulation also increases opportunities for strategic doubt. [JSTOR]jstor.orgDeep Fakes: A Looming Challenge for Privacy, Democracy…by B Chesney · 2019 · Cited by 2856 — Hence what we call the liar's divide…
This dynamic is closely connected to broader trends in declining trust, political polarisation, and accusations of “fake news”. AI-generated media adds another layer by providing a technically credible explanation for dismissing inconvenient evidence. [Cambridge University Press & Assessment]cambridge.orgdeepfakes, leading to what Chesney and Citron (Reference Chesney and Citron2019) term the “liar's dividend.” While this tactic of denial…
How synthetic media weakens genuine evidence
The liar’s dividend does not require viewers to believe a specific alternative story. Often it is enough to create uncertainty.
Doubt can be more powerful than persuasion
Traditional misinformation usually tries to convince people that a false claim is true. The liar’s dividend works differently. A person may not need others to accept a replacement explanation. They only need enough people to think, “Maybe the image was AI-generated.”
That uncertainty can delay accountability, complicate investigations, and fragment public understanding of events. UNESCO has described this broader effect as part of a “crisis of knowing”, in which authentic recordings can be dismissed simply because synthetic alternatives are known to exist. [UNESCO]unesco.orgdeepfakes and crisis knowingDeepfakes and the crisis of knowing27 Oct 2025 — Deepfakes and the crisis of knowing. As deepfakes blur reality… The liar's divi…
Audiences become more sceptical of authentic media
Recent research suggests that the growth of synthetic media may produce a general scepticism shift. In a large-scale 2026 study on audio deepfakes, researchers found that people were not dramatically worse at identifying fake audio than before, but they had become significantly less likely to trust genuine recordings. The central problem was increasing distrust of authentic content rather than improved deception alone. [arXiv]arxiv.orgEroding Trust in Real Speech: A Large-Scale Study of Human Audio Deepfake PerceptionMay 21, 2026…
Although the study focused on audio, the underlying concern applies to visual evidence as well. If people assume that convincing fabrication is easy, they may begin treating authentic photographs and videos with greater suspicion.
Real images can be misidentified as AI-generated
The information environment becomes even more confusing when genuine images are incorrectly labelled as AI creations. Analysis of online political content has found widespread instances of users misidentifying authentic images as AI-generated, illustrating how suspicion can spread beyond genuinely manipulated media. [Financial Times]ft.comFinancial Times How we were deepfaked by election deepfakesDuring various global elections, several deepfakes like Joe Biden's robocalls and false videos of Marine Le Pen’s family appeared. Nevert…
In such cases, the damage comes not from successful forgery but from eroded confidence in reality-based evidence.
Why some denials succeed
Not every claim of “that image is AI-generated” is persuasive. Several factors influence whether a denial gains traction.
Existing distrust
Denials are more likely to work when audiences already distrust institutions, journalists, political opponents, or other sources of information. In these environments, claims that evidence was fabricated may fit pre-existing beliefs. Research on the liar’s dividend suggests that strategic accusations of fakery can help public figures maintain support among sympathetic audiences. [IDEAS/RePEc]ideas.repec.orgIDEAS/RePEcThe Liar's Dividend: The Impact of Deepfakes and Fake…by KJ Schiff · 2023 · Cited by 13 — This study examines the phenomeno…
Technical complexity
Most people cannot independently verify whether a photograph or video is authentic. Digital forensics, metadata analysis, provenance systems, and source verification often require specialist expertise. This knowledge gap creates space for competing claims about authenticity. [Thomson Reuters]thomsonreuters.comThomson ReutersDeepfakes on trial: How judges are navigating AI evidence…May 8, 2025 — “I think the courts will see [the liar's divide…
Speed of information sharing
Images and videos often spread much faster than verification efforts. A denial can circulate immediately, while authentication may take days or weeks. During that period, uncertainty itself can become influential.
What makes a denial credible or weak
Because AI-generated media is real and increasingly capable, dismissing concerns outright is not reasonable. The key question is whether a denial is supported by evidence.
A denial becomes more credible when it is accompanied by:
- Independent forensic analysis.
- Evidence of manipulation or editing.
- Missing or inconsistent provenance records.
- Corroborating technical findings from multiple investigators.
- Demonstrable flaws in the visual material itself. [Thomson Reuters]thomsonreuters.comThomson ReutersDeepfakes on trial: How judges are navigating AI evidence…May 8, 2025 — “I think the courts will see [the liar's divide…
A denial becomes weaker when it relies primarily on assertion:
- Claims that “anything can be faked nowadays”.
- No supporting forensic evidence.
- Rejection of multiple independent verifications.
- Contradiction by eyewitnesses, original files, or source documentation.
- Shifting explanations about why the evidence is supposedly false. [Thomson Reuters]thomsonreuters.comThomson ReutersDeepfakes on trial: How judges are navigating AI evidence…May 8, 2025 — “I think the courts will see [the liar's divide…
Courts, journalists, and investigators increasingly focus on authentication procedures rather than visual appearance alone. Questions about chain of custody, original files, timestamps, and corroborating records become more important when visual realism can no longer be taken for granted. [Thomson Reuters]thomsonreuters.comThomson ReutersDeepfakes on trial: How judges are navigating AI evidence…May 8, 2025 — “I think the courts will see [the liar's divide…
The challenge for information integrity
The liar’s dividend highlights a crucial point about artificial intelligence and information integrity: the greatest risk is not always that people believe false images. Sometimes the greater risk is that people stop believing true ones.
In this sense, synthetic media changes the role of visual evidence itself. Photographs and videos no longer function as automatic proof in the way many people once assumed. Authenticity increasingly depends on context, provenance, corroboration, and verification rather than appearance alone. As AI-generated images become more realistic, protecting trust in genuine evidence may become as important as detecting fabricated content. [UNESCO+2California Law Review]unesco.orgdeepfakes and crisis knowingDeepfakes and the crisis of knowing27 Oct 2025 — Deepfakes and the crisis of knowing. As deepfakes blur reality… The liar's divi…
Amazon book picks
Further Reading
Books and field guides related to When Real Images Become Easier to Deny. Use these as the next step if you want deeper reading beyond the article.
A History of Fake Things on the Internet
Examines how fake and real content become harder to distinguish.
eBay marketplace picks
Marketplace Samples
Example marketplace items related to this page. Use the search link to explore similar finds on eBay.
Endnotes
-
Source: jstor.org
Link: https://www.jstor.org/stable/26891938Source snippet
Deep Fakes: A Looming Challenge for Privacy, Democracy...by B Chesney · 2019 · Cited by 2856 — Hence what we call the liar's divide...
-
Source: cambridge.org
Link: https://www.cambridge.org/core/journals/american-political-science-review/article/liars-dividend-can-politicians-claim-misinformation-to-evade-accountability/687FEE54DBD7ED0C96D72B26606AA073Source snippet
deepfakes, leading to what Chesney and Citron (Reference Chesney and Citron2019) term the “liar's dividend.” While this tactic of denial...
-
Source: ideas.repec.org
Link: https://ideas.repec.org/p/osf/socarx/x43ph.htmlSource snippet
IDEAS/RePEcThe Liar's Dividend: The Impact of Deepfakes and Fake...by KJ Schiff · 2023 · Cited by 13 — This study examines the phenomeno...
-
Source: unesco.org
Title: deepfakes and crisis knowing
Link: https://www.unesco.org/en/articles/deepfakes-and-crisis-knowingSource snippet
Deepfakes and the crisis of knowing27 Oct 2025 — Deepfakes and the crisis of knowing. As deepfakes blur reality... The liar's divi...
-
Source: arxiv.org
Link: https://arxiv.org/abs/2605.26136Source snippet
Eroding Trust in Real Speech: A Large-Scale Study of Human Audio Deepfake PerceptionMay 21, 2026...
Published: May 21, 2026
-
Source: californialawreview.org
Link: https://www.californialawreview.org/print/deep-fakes-a-looming-challenge-for-privacy-democracy-and-national-securitySource snippet
California Law ReviewDeep Fakes: A Looming Challenge for Privacy, Democracy...And, in what we call the “liar's dividend,” deep fakes mak...
-
Source: ft.com
Title: Financial Times How we were deepfaked by election deepfakes
Link: https://www.ft.com/content/62d81e6c-eec0-4d09-a71f-6aba579912ddSource snippet
During various global elections, several deepfakes like Joe Biden's robocalls and false videos of Marine Le Pen’s family appeared. Nevert...
-
Source: thomsonreuters.com
Link: https://www.thomsonreuters.com/en-us/posts/ai-in-courts/deepfakes-evidence-authentication/Source snippet
Thomson ReutersDeepfakes on trial: How judges are navigating AI evidence...May 8, 2025 — “I think the courts will see [the liar's divide...
Published: May 8, 2025
-
Source: thomsonreuters.com
Title: ai evidence trials
Link: https://www.thomsonreuters.com/en-us/posts/ai-in-courts/ai-evidence-trials/Source snippet
AI evidence in jury trials: Navigating the new frontier of justice6 Oct 2025 — Unacknowledged AI-generated evidence, on the other hand, c...
-
Source: Wikipedia
Link: https://en.wikipedia.org/wiki/ReutersSource snippet
ReutersReuters is a news agency wholly owned by Thomson Reuters, a multinational information conglomerate. It employs around 2,500 jou...
-
Source: thomsonreuters.com
Link: https://www.thomsonreuters.com/en-us/posts/technology/practice-innovations-deepfakes/Source snippet
Seeing is no longer believing — the rise of deepfakesJul 18, 2023 — Practice Innovations: Seeing is no longer believing — the rise of dee...
-
Source: osf.io
Link: https://osf.io/download/q6mwnSource snippet
1 allows for examination of the Deepfakes Hypothesis as well as the Liar's Dividend Hypoth- esis through exploring...Read more...
-
Source: threads.com
Title: unesco org en articles deepfakes and crisis knowing unesco on the liars
Link: https://www.threads.com/%40drmurji/post/DW5bo76jfV-/unesco-org-en-articles-deepfakes-and-crisis-knowing-unesco-on-the-liarsSource snippet
Deepfakes and the crisis of knowing. unesco.org. Deepfakes and the crisis of knowing. 2.Read more...
Additional References
-
Source: aicerts.ai
Link: https://www.aicerts.ai/news/synthetic-images-and-the-objective-reality-crisis/Source snippet
Synthetic Images and the Objective Reality CrisisFurthermore, the "liar's dividend" allows guilty actors to dismiss authentic photos as s...
-
Source: medium.com
Link: https://medium.com/%40tahindukajoseph/the-liars-dividend-reflections-on-a-global-crisis-of-trust-daa8748848fbSource snippet
The Liar's Dividend: Reflections on a Global Crisis of TrustIn a world where deepfakes and misinformation can be produced with a few clic...
-
Source: mondaq.com
Link: https://www.mondaq.com/india/fin-tech/1730990/when-seeing-is-no-longer-believing-deepfakes-the-liars-dividend-and-the-crisis-of-digital-evidence-in-the-legal-systemSource snippet
When Seeing Is No Longer Believing: Deepfakes, The...15 Jan 2026 — The "Deepfake Defence" and the resulting Liar's pidend risk creating...
-
Source: biometricupdate.com
Link: https://www.biometricupdate.com/202508/the-liars-dividend-deepfakes-synthetic-media-and-the-cybersecurity-disinformation-crisisSource snippet
The Liar's Dividend: Deepfakes, synthetic media, and...6 Aug 2025 — The Liar's Dividend: Deepfakes, synthetic media, and the cybersecuri...
-
Source: knightcolumbia.org
Link: https://knightcolumbia.org/content/dont-panic-yet-assessing-the-evidence-and-discourse-around-[generative-aiSource snippet
deepfake now takes minutes rather than days, many propagandists... Meanwhile, the risk of a 'liar's dividend'—where politicians or...Re...
-
Source: ubaltlawreview.com
Link: https://ubaltlawreview.com/2025/12/01/deepfakes-in-the-courtroom-challenges-in-authenticating-evidence-and-jury-evaluation/Source snippet
University of Baltimore Law ReviewDeepfakes in the Courtroom: Challenges in Authenticating...Dec 1, 2025 — Deepfakes in the Courtroom: C...
-
Source: theinnovator.news
Link: https://theinnovator.news/ai-powered-disinformation-and-the-liars-dividend/Source snippet
2 years ago. by Kay... deepfakes. Although the two issues are often linked, disinformation deserves...Read more...
-
Source: gnet-research.org
Title: from confusion to extremism how deepfakes facilitate radicalisation
Link: https://gnet-research.org/2026/02/11/from-confusion-to-extremism-how-deepfakes-facilitate-radicalisation/Source snippet
From Confusion to Extremism: How Deepfakes Facilitate...11 Feb 2026 — At the same time, Dr Naffi identifies that deepfakes create a liar...
-
Source: researchgate.net
Link: https://www.researchgate.net/publication/393070921_The_Liar%27s_Dividend_How_Disinformation_Erodes_Trust_and_Shields_DeceitSource snippet
it public awareness of disinformation and deepfakes to dismiss real evidence as...Read more...
-
Source: brookings.edu
Title: Watch out for false claims of deepfakes, and actual
Link: https://www.brookings.edu/articles/watch-out-for-false-claims-of-deepfakes-and-actual-deepfakes-this-election-year/Source snippet
liar's dividend originated out of concerns about deepfakes. Still, our most recent study does suggest that as awareness of deepfake image...
Topic Tree