You can learn a lot about a group of people by watching what it refuses to believe in.
Right now, the American electorate is entirely unwilling to entertain the legitimacy of an election they lose.
Ahead of the 2026 midterms, the public narrative is hyper-fixated on generic ballot polling and congressional maps. But a new survey of more than 2,000 U.S. adults looks at a much more volatile asset: institutional trust.
Key Takeaways
- Trust is conditional: 92.5% of Democrats would trust a Democratic House victory, but only 34.3% would trust a Republican win.
- The youth struggle: 53.0% of 18-to-24-year-olds say misinformation has made choosing their vote harder, compared to just 27.8% of voters 65+.
- AI as the new fact-checker: One in three respondents (33.8%) have used AI for midterm campaign information.
- The authenticity crisis: More than a third (36.7%) report previously mistaking AI-generated political content for genuine material.
- Dishonesty is expected: Republicans (17.1%) and Democrats (9.4%) show a measurable willingness to consider candidates who deliberately spread false information.
The data reveals an electorate that is drowning in conflicting narratives, outsourcing its political vetting to algorithms, and deciding what is true based largely on the outcome they prefer.
We like to talk about trust in elections as a baseline civic assumption. The survey data suggests it is actually in very short supply in 2026.
Enjoying Covers content? Add us as a preferred source on your Google account
The Conditional Confidence Gap
If you want to know whether a voter trusts the electoral process, don't ask them to read the Constitution. Just ask them who they think is going to win.
When presented with a hypothetical Democratic House victory, a commanding 92.5% of Democrats report they would completely or mostly trust the result. But if Republicans win the House? That trust collapses to 34.3%.
The right shows a similar, though slightly less severe, elasticity.
| Political Affiliation | Trust Democrat Victory | Trust Republican Victory |
|---|---|---|
| 🔵 Democrats | 92.5% | 34.3% |
| 🔴 Republicans | 45.9% | 83.5% |
| ⚪️ Independents | 70.3% | 44.7% |
Collapse the negative sentiment, and the vulnerability of the system becomes glaring. Three in five Democrats (60.7%) report low or no trust in a Republican House victory. Half of Republicans (50.0%) report low or no trust in a Democratic victory.
Low trust does not strictly mean these voters will refuse to accept the results, but it proves that institutional confidence isn't standard issue: it is the payout for holding the winning ticket.
The misinformation game
Everyone agrees campaign misinformation is an obvious liability. They just disagree on who is manufacturing it.
Across the full sample, 57.2% of respondents blame Republicans for spreading more campaign misinformation, while 17.2% blame Democrats, and 20.8% blame both equally. The independent vote, the demographic that actually swings midterms, shows its hand clearly: 49.6% of Independents blame the GOP, compared to just 11.8% who blame Democrats.
But identifying a liar and punishing one are two different mechanics. When asked if they would still consider a candidate who deliberately spread false information, the survey reveals that dishonesty is expected as part of the election cycle.
| Political Affiliation | Would Still Consider Voting for Them |
|---|---|
| 🔴 Republicans | 17.1% |
| 🔵 Democrats | 9.4% |
| ⚪️ Independents | 5.7% |
Independents prove the most rigid, with only 5.7% willing to overlook deliberate deception. For the partisans, the threshold for disqualification is surprisingly flexible.
Digital Natives Are Drowning in the Noise
The conventional media narrative assumes older voters are the most susceptible to online noise. The survey data says the exact opposite.
Overall, 39.1% of respondents say false or misleading information has made deciding who to vote for harder. But look at the generational splits. The difficulty strictly decreases across every single age band.
- 18–24: 53.0%
- 25–34: 45.1%
- 35–44: 42.3%
- 45–54: 36.4%
- 55–64: 34.1%
- 65+: 27.8%
Gen Z isn't navigating the digital landscape with ease; they are suffocating in it. More than half of young adults find the baseline task of choosing a candidate harder because of the information environment.
Separately, 20.4% of 18-to-24-year-olds actually report sharing political information online that they later discovered was false, compared to just 8.2% of the 65-and-older demographic.
Outsourcing Truth to the Machine
If the information ecosystem is broken, voters are looking for a new referee. For a measurable slice of the electorate, that referee is artificial intelligence.
During this midterm cycle, 33.8% of respondents have used AI for at least one listed political task. They aren't just asking chatbots where their polling place is (8.2%). They are actively using the technology to vet the political landscape:
- 21.5% used AI to check whether a political claim was true.
- 19.5% used it to find candidate or policy information.
- 12.1% used it to compare candidates or parties.
- 3.5% requested a candidate or party recommendation.
Demographically, men are notably leaning on this technology more than women (39.7% vs. 28.5% for any listed AI use).
Here is the structural irony of the 2026 midterms: Voters are using AI to find the truth, while simultaneously being deceived by it. More than a third of respondents (36.7%) report mistaking political AI content for genuine material in the past.
When asked to select up to two specific concerns regarding AI content, respondents pointed heavily toward fabricated events (63.1%) and false policy claims (61.2%), rather than fake celebrity endorsements (8.8%).
The electorate is deploying algorithms to protect themselves from algorithmic deception.
Economic Noise and Local Radars
When it comes to the specific topics where voters suspect they are being misled, the economy dominates. A full 72.8% of respondents report suspected misinformation regarding the economy, prices, and jobs. This comfortably outpaces concerns about candidates’ personal lives (59.0%) and immigration (56.2%).
State-by-state survey findings
| State | Expected Midterms Outcome | Comments |
| Arizona | 🔵 Dem Governor (91%) 🔴 Rep House toss-up district (AZ-02 Lean R 69%) |
Only 3.5% report ever sharing political information later found misleading (4th lowest nationally); 52.6% suspect misinformation about crime and public safety. |
| California | 🔵 Dem Governor (97%) | 39.5% report using AI for midterm information (9th highest); 60.5% suspect misinformation about voting and election results. |
| Connecticut | 🔵 Dem Governor (96%) | Partisan trust divides sharply: 91.3% would trust a Democratic House win (2nd highest), versus 26.1% a Republican win (2nd lowest); 56.5% use AI for midterm tasks. |
| Florida | 🔴 Rep Senate (89%) & Rep Governor (71%) 🔵/🔴 Split House races (FL-14/22 Lean D, FL-07/09/13/27 Lean R) |
39.4% have used AI for midterm information; 17.6% used it to compare candidates or parties (10th highest). |
| Georgia | 🔵 Dem Senate (96%) & Dem Governor (61%) | At 17.6%, respondents are twice as likely as the national average (8.8%) to cite fake celebrity endorsements as a top AI concern (6th highest). |
| Iowa | 🟡/🔴 Toss-up Senate (R 53%) 🔵 Dem Governor (83%) & Dem House toss-up (IA-02 Lean D 72%) |
20.0% would still consider voting for a preferred candidate who deliberately spread false information (2nd highest nationally); 50.0% say misinformation made their choice harder. |
| Kansas | 🟡/🔴 Toss-up Senate (R 52%) 🔴 Rep Governor (73%) |
Leads in immigration-related misinformation concerns: 81.0% suspect misinformation about immigration and border security (highest nationally). |
| Maryland | 🔵 Dem Governor (97%) | Ranks 1st for misinformation-related voter confusion: 60.5% say it made choosing whom to vote for harder; 21.1% used AI to find out how, where, or when to vote (2nd highest). |
| Michigan | 🔵 Dem Senate (71%) & Dem Governor (92%) 🔵 Dem House toss-up district (MI-07 Lean D 74%) |
17.5% would still consider a candidate who deliberately spread false information (4th highest); 70.2% cite candidate character as the most suspected misinformation area. |
| Minnesota | 🔵 Dem Senate (90%) & Dem Governor (97%) 🔴 Rep House toss-up district (MN-01 Lean R 63%) |
Candidate character outpaces economic issues as a suspected source of falsehoods: 78.1% suspect misinformation about candidate character (2nd highest nationally). |
| Nevada | 🔵 Dem Governor (78%) 🔴 Rep House toss-up district (NV-02 Toss-up R 59%) |
76.0% suspect misinformation about candidates' character (4th highest); 64.0% have low or no trust in a Republican House win (5th highest). |
| New Jersey | 🔵 Dem Senate (97%) 🔴 Rep House toss-up district (NJ-02 Lean R 72%) |
48.0% have mistaken AI-generated political content for genuine material (7th highest nationally; highest among larger state samples). |
| New York | 🔵 Dem Governor (96%) 🔵 Dem House toss-up district (NY-17 Lean D 63%) |
Has the lowest share saying misinformation made choosing a candidate harder (23.9%); 71.6% flag fabricated event imagery as a top AI concern (7th highest). |
| North Carolina | 🔵 Dem Senate (97%) 🔴 Rep House toss-up districts (NC-03 Toss-up R 60%; NC-09/13 Lean R) |
Only 24.0% say misinformation complicated vote choice (2nd lowest); 0.0% asked AI to recommend a candidate or party. |
| Ohio | 🔵 Dem Senate (68%) 🔵 Dem Governor (81%) 🔴 Rep House toss-up districts (OH-10/15 Lean R) |
17.1% would consider a candidate who deliberately spreads false claims (6th highest); 53.9% suspect misinformation about healthcare and public health (also 6th highest). |
| Pennsylvania | 🔵 Dem Governor (98%) 🔴 Rep House toss-up district (PA-01 Toss-up R 53%) |
16.7% would still consider a candidate caught deliberately spreading false information (joint 8th highest); only 26.9% say misinformation made choosing a candidate harder (5th lowest). |
| South Carolina | 🔴 Rep Senate (71%) & Rep Governor (90%) 🔴 Rep House toss-up district (SC-01 Lean R 63%) |
Shows the strongest repeated AI-adoption pattern nationally: 54.3% use AI for midterm information (4th highest), 40.0% for fact-checking (4th), and 31.4% to compare candidates (2nd). |
| Texas | 🔵 Dem Senate (69%) 🔴 Rep Governor (64%) 🔵/🔴 Split House toss-up districts (TX-35 Lean D; TX-23 Toss-up R) |
25.1% actively use AI to fact-check political claims (10th highest nationally); 52.7% report suspected misinformation about crime and public safety. |
| Virginia | 🔵 Dem Senate (99%) 🔵 Dem House toss-up district (VA-01 Lean D 64%) |
Trust differs by 54.5 points: 81.8% would trust a Democratic House win versus 27.3% a Republican win (3rd-lowest Republican-win trust nationally). |
| Wisconsin | 🔵 Dem Governor (84%) 🔴 Rep House toss-up district (WI-01 Toss-up R 54%) |
48.8% name fake candidate speeches as a top AI concern (2nd highest nationally); 80.5% suspect misinformation about economic reporting. |
Note: Any states with less than five respondents in the survey were excluded from the state-by-state analysis
Methodology
This survey consists of over 2,000 completed submissions from U.S. adults, surveyed in October 2026. Results are unweighted and descriptive, measuring perceptions and reported behaviors rather than independently verified instances of misinformation exposure.
Political affiliation reflects self-reported identity, not actual vote choice, and national or state representativeness is not established.
Publication status is provisional pending the resolution of 19 submissions containing substantive multiple selections alongside neutral answers, and figures will be refreshed before final publication. No statistical significance or causal claims are made.
