An AI-generated image of Donald Trump in a Pittsburgh Steelers jersey did not need to convince viewers that he was literally a football player to have political meaning. It could connect him with local sports culture and make supporters feel that he belonged to their community. That distinction—between correcting a false claim and addressing the feeling or identity an image conveys—was the point of University of Washington researcher Danielle Lee Tomson’s November 2024 conversation with GeekWire.
What Tomson meant by “you can’t fact-check a feeling”
GeekWire published its interview with Tomson, then research manager for election rumors at the University of Washington’s Center for an Informed Public (CIP), on November 2, 2024, three days before the U.S. election. The conversation was about a political effect that does not depend on a viewer mistaking synthetic content for documentary evidence: an image can be obviously artificial yet still evoke affiliation, humor, patriotism, grievance, or belonging. Read the GeekWire interview.
There are several distinct questions to ask about political content:
- Factual accuracy: Does the image, recording, or claim represent what actually happened?
- Emotional meaning: What does it make viewers feel about a candidate, group, or political moment?
- Social signaling: What does sharing it say to other people—“I’m one of you,” “this is funny,” or “I’m angry too”?
- Ambient persuasion: What impression accumulates through repeated exposure to a stream of content, rather than a single explicit argument?
A fact-check can establish that a picture is synthetic or that an event did not occur. That correction may not remove the humor, affinity, or group meaning that made someone share it. The factual question still matters; it simply does not exhaust the political one.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →#1 Best Overall
Why the story was more than deepfakes
Much public concern about AI in the 2024 campaign centered on convincing fake videos, impersonated candidate audio, fabricated images, and cheap, scalable propaganda. Those risks are real, but Tomson’s “vibe” point concerns a different use: content that does not have to pass as authentic to reinforce a political identity or mood.
AI-generated campaign-era material can function as a meme, satire, political fan art, aspirational imagery, or an identity marker. A fake image can be politically consequential without being a successful factual hoax. Conversely, its circulation does not by itself prove that viewers believed it, were persuaded by it, or changed their vote. GeekWire’s interview offers a conceptual analysis and examples, not a representative study measuring exposure or electoral effect.
Three ways AI can affect political information
Deception: fabricated evidence or instructions
A synthetic recording presented as genuine can mislead people about what a candidate said or did. The GeekWire conversation contrasted expressive imagery with an AI-generated voice robocall that used a candidate’s voice to give voters the wrong day and location for voting. That is not merely atmosphere: false practical instructions could interfere with voting and raise legal consequences. It calls for rapid verification and authoritative election information, not just interpretation of the image’s symbolism. The interview discusses both examples.
Amplification: making repetition look like many voices
AI can vary the wording of a repeated claim. When related messages appear in different language, readers may mistake repetition for independent confirmation. A University of Washington experiment provides evidence for this mechanism under study conditions; it does not show that every AI-generated variation has that effect in real-world campaigning.
Free tools Windows power users keep installed
One-click scans. No signup required.
Rank #3
Atmosphere: reinforcing identity and emotion
Symbolic imagery and memes can attach a candidate to a place, community, or cultural reference. Their function may be to express solidarity or create a shared political mood rather than to assert a verifiable fact. Determining the creator’s intent, audience, and distribution requires context; a synthetic image alone does not establish that it was official campaign material or propaganda.
Why UW researchers use the word “rumor”
CIP uses “rumor” broadly for claims circulating as people try to make sense of events. A rumor may be true, false, partly true, unsubstantiated, or misleading because of selective framing. The term avoids deciding too early whether a claim is misinformation, disinformation, or something more complicated—especially when intent is hard to establish. The UW Information School describes the election-rumor work.
This approach also helps explain why a political narrative can combine recognizable facts with invented context, exaggeration, or symbolic imagery. CIP’s election-rumor project monitored how communities discussed candidates, voting procedures, and election-related claims across online and offline settings. Its research overview describes this broader work on how people interpret information and events. See CIP’s research overview.
What fact-checking can—and cannot—establish
Fact-checking is most direct when a claim is specific, attributable, time-bounded, and testable against evidence. It can establish that a purported voting location is wrong, a recording is fabricated, or a claim lacks support. It is less able, on its own, to answer questions such as whether an image made someone feel represented or whether a meme expressed a group’s shared anger.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →That is a limitation of scope, not a reason to abandon verification. Corrections can address factual claims; they cannot automatically undo identity, trust, repetition, or social meaning. Nor does a share prove belief: people circulate material for humor, irony, outrage, or affiliation as well as persuasion.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What the later UW experiment adds
A preregistered experiment published September 10, 2024, included 1,200 U.S. participants and tested AI-paraphrased versions of repetitive disinformation messages. The researchers reported that the AI-paraphrased messages increased perceived social consensus for false claims, particularly among participants less familiar with the false narrative. Among Republican participants in the study, the messages also increased belief in false claims compared with the control condition. The study additionally reported greater sharing intentions and recall of broader false claims. The Center for Statistics and the Social Sciences summarizes the experiment.
Those findings concern AI-assisted variation and repetition, not the emotional effect of a particular synthetic image. They are experimental findings, not a direct measurement of what happened across the electorate. Neither the study nor Tomson’s pre-election interview establishes that AI determined the 2024 result. Later CIP work describes rapid, interdisciplinary monitoring of how people make sense of elections and crises, rather than a single causal verdict about an election outcome. Read the UW repository document.
How to assess an AI political post before sharing it
- Ask what kind of claim it makes. Is it offered as a factual record, a joke, an expression of support, or a practical instruction?
- Check provenance. Look for the original source, date, and context. Do not infer that an image is an official campaign asset just because it features a candidate.
- Identify the emotional cue. Consider whether it is trying to evoke pride, outrage, fear, humor, or belonging. That can matter even if you recognize that it is synthetic.
- Watch for repetition in new wording. Several versions of a claim are not necessarily several independent sources.
- Verify urgent voting information with election authorities. Do not rely on a viral recording or meme for dates, locations, or procedures.
- Separate spread from persuasion. A post’s popularity or your own impulse to share it does not show that its factual claim is true or that other people believe it.
Where the evidence ends
Tomson’s interview is a timely pre-election analysis of how synthetic media can work through feeling and identity, illustrated with selected examples. It does not quantify how many voters encountered those examples, establish whether they believed them, compare their effects systematically with conventional memes or campaign advertising, or prove that they changed the election result. The later UW experiment adds evidence about AI-assisted repetition in a controlled study, but answers a different question.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
The useful takeaway is narrower and stronger than a claim that AI either won or failed to win an election: political media can matter without persuading viewers that a fabrication is real. Understanding its effects requires checking factual claims while also asking what emotions, identities, and social bonds the content may reinforce.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
