Electoral Impact Assessment Model

Overview

AI dramatically lowers the cost of creating and distributing disinformationRiskAI DisinformationPost-2024 analysis shows AI disinformation had limited immediate electoral impact (cheap fakes used 7x more than AI content), but creates concerning long-term epistemic erosion with 82% higher beli...Quality: 54/100 at scale. But does this translate to meaningful impact on election outcomes? This model provides a framework for estimating the marginal effect of AI-generated disinformation on electoral results and democratic processes.

Core Question: By how much can AI disinformation shift election results, and under what conditions?

Strategic Importance

Understanding AI disinformation's electoral impact matters because democratic legitimacy depends on elections reflecting genuine voter preferences. If AI disinformation can reliably shift 2-5% of votes in close elections (our central estimate), this represents a fundamental threat to democratic governance.

Magnitude Assessment

Resource Implications

Key Cruxes

Key insight: The marginal impact of AI disinformation is probably smaller than media coverage suggests for individual elections, but systemic effects on democratic trust may matter more than vote margin shifts.

Parameter Estimates

The following table summarizes key model parameters derived from empirical research and expert elicitation.

Research from MIT Sloan found that false information spreads 70% faster than true information on social media, with political falsehoods showing particularly rapid diffusion. This suggests AI-generated disinformation may benefit from inherent platform dynamics that favor novel, emotionally engaging content.

The Marginal Impact Problem

Elections are influenced by countless factors:

• Economic conditions
• Candidate quality
• Campaign spending
• Media coverage
• Debates and events
• Ground operations
• Traditional advertising
• Disinformation (pre-AI)
• AI-Generated disinformation (new)

Challenge: Isolating the marginal contribution of AI-enhanced disinformation from everything else.

Impact Pathway Model

We can decompose the causal pathway from AI capability to electoral impact:

flowchart TD
  AI[AI Capability<br/>Quality improvement: 30%/yr] --> DISINFO[Disinformation Volume/Quality<br/>150-3000x increase]
  DISINFO --> EXPOSE[Audience Exposure<br/>1.5-4x multiplier vs traditional]
  EXPOSE --> BELIEF[Belief Change<br/>2-6x multiplier for AI content]
  BELIEF --> VOTE[Vote Choice Change<br/>5-15% of exposed shift]
  VOTE --> OUTCOME[Election Outcome Change<br/>Critical in races within 3%]

  style AI fill:#ffddcc
  style OUTCOME fill:#ff9999

Each step has a probability/magnitude. The overall impact is the product of all steps.

Step 1: AI → Disinformation Volume/Quality

Pre-AI Disinformation Constraints:

• Human effort required for each piece of content
• Limited personalization
• Detectable patterns (template-based)
• Cost: $1-10 per piece for quality content

AI Enhancement:

• Automated generation at massive scale
• Personalized to individual targets
• High quality, indistinguishable from organic content
• Cost: $0.001-0.01 per piece

Multiplier Effect:

• Volume increase: 100-1000x
• Quality increase: 1.5-3x (more convincing)
• Personalization increase: 10-100x (targeted messaging)

Overall AI Impact on Content Creation: ~150-3000x increase in effective disinformation output

Confidence: High. Well-documented in 2024 elections.

Step 2: Volume/Quality → Exposure

Not all content reaches audiences. Social media algorithms, platform moderation, and user behavior filter content.

Platform Moderation:

• Platforms remove ~20-40% of detected disinformation
• AI-generated content currently detected at ~30-60% rate (falling)
• Net effect: 50-80% of AI disinformation reaches audiences (vs ~60-90% of human disinformation)

Algorithmic Amplification:

• Engaging content (often outrage-inducing disinformation) promoted
• AI-generated content can optimize for engagement
• Multiplier: 1.2-2x amplification vs. baseline

Audience Reach:

• Traditional disinformation: reaches 5-15% of target audience
• AI-personalized disinformation: reaches 10-30% of target audience (better targeting)

Overall Exposure Multiplier (AI vs traditional): 1.5-4x

Confidence: Medium. Platform algorithms are opaque; estimates based on disclosed data.

Step 3: Exposure → Belief Change

How many people who see disinformation actually believe it?

Baseline Belief Rates (Pre-AI):

• Aligned with existing beliefs: 30-50% believe
• Counter to existing beliefs: 5-15% believe
• No prior opinion: 20-40% believe

AI Enhancement Factors:

Personalization: AI can tailor messaging to individual psychology

• Estimated increase in persuasiveness: 1.3-2x

Multimodal Content: Deepfakes, voice clones more convincing than text

• Estimated increase for video/audio: 1.5-2.5x vs text

Repetition at Scale: Multiple exposures via different "sources" (all AI)

• Estimated increase per additional exposure: 1.2x (up to 3-4 exposures)

Overall Belief Change Multiplier (AI vs traditional): 2-6x depending on content type and targeting

Confidence: Low-Medium. Limited experimental data. Based on persuasion research and preliminary studies.

Step 4: Belief Change → Vote Choice Change

Not all belief changes translate to vote switching.

Baseline Vote Impact (pre-AI disinformation):

• Partisans rarely switch: 1-3% affected
• Swing voters more susceptible: 10-20% affected
• Low-information voters most susceptible: 15-30% affected

Election Type Matters:

• Presidential elections: Voters have strong priors, hard to shift
• Local elections: Lower information, easier to influence
• Ballot initiatives: Voters often uncertain, highly influenceable

AI Disinformation Vote Impact: Assuming AI increases belief change by 2-6x (Step 3):

• Partisans: 2-8% affected (low end—beliefs don't translate to switching)
• Swing voters: 15-35% affected
• Low-info voters: 25-50% affected

Weighted Average (typical electorate):

• ~15% swing voters
• ~30% low-info voters
• ~55% strong partisans

Overall Vote Impact: 5-15% of exposed population might shift vote due to AI disinformation

Confidence: Low. Vote switching is multi-causal; attribution difficult.

Step 5: Vote Change → Outcome Change

Finally, how many votes need to shift to change election results?

Close Elections:

• 2020 U.S. Presidential: Decided by ~44,000 votes across 3 states (~0.03% of total votes)
• Many congressional races decided by 1-3%
• Close elections highly vulnerable to small shifts

Landslide Elections:

• 10+ point margins require massive shifts to overturn
• AI disinformation unlikely to swing

Quantitative Model:

Assume:

• Close election (within 3%)
• AI disinformation reaches 30% of electorate
• Of those, 10% shift votes
• Overall vote shift: 3%

Result: Enough to flip a close election.

Case Study Analysis

2024 Elections: The "AI Election" That Wasn't?

Despite being called the "AI election year," post-election analysis found limited evidence of decisive AI disinformation impact.

Why the limited impact?

Possible Explanations:

1. Detection Worked: Platform moderation caught enough AI content to limit spread

   • Evidence: Multiple platforms reported removing AI-generated campaigns
   • Counter-evidence: Much went undetected

2. Audience Skepticism: Voters increasingly aware of AI manipulation, more skeptical

   • Evidence: Increased media literacy campaigns
   • Counter-evidence: Most voters unaware of specific AI threats

3. Cheap Fakes More Effective: Simple edited videos outperformed sophisticated AI (7:1 ratio per News Literacy Project)

   • Evidence: Well-documented
   • Implication: Quality may matter less than simplicity

4. Existing Polarization Dominates: Voters already so polarized that marginal disinformation doesn't matter

   • Evidence: Historically high partisan loyalty
   • Implication: AI disinformation adds noise, not signal

5. Measurement Problem: Impact exists but is undetectable amid other factors

   • Evidence: Close races in swing states consistent with small AI impact
   • Problem: Can't prove counterfactual

Most Likely: Combination of #3, #4, and #5. AI disinformation had some impact but was not decisive in 2024.

Romania 2024: First Election Annulled Over AI Interference

Event: Romania's Constitutional Court annulled the results of the first round of its presidential election on December 6, 2024 — the first time in any democracy that election results were voided specifically due to evidence of AI-powered foreign interference.

Background: Călin Georgescu, a far-right, pro-Russia candidate polling at under 5%, surged to win the first round with 22.9% of the vote. Romanian intelligence services (SRI, SIE, STS) subsequently declassified evidence showing a coordinated campaign involving:

• TikTok manipulation: Over 25,000 TikTok accounts activated simultaneously to promote Georgescu, many dormant accounts that suddenly became active. A Global Witness investigation found systematic amplification through the platform's algorithm.
• AI-generated content: AI tools used to create and optimize campaign messaging, with content designed to evade platform detection.
• Undisclosed foreign funding: Approximately $1 million in campaign spending routed through cryptocurrency and foreign accounts, violating Romanian campaign finance law.
• Coordinated inauthentic behavior: Networks of accounts posting synchronized content across platforms, consistent with state-backed influence operations. Romanian intelligence attributed the campaign to Russian-linked actors.

Constitutional Court ruling: The court found that the electoral process had been "vitiated" and that "the equality of chances of the candidates" had been compromised. The ruling cited both the undisclosed financing and the coordinated social media manipulation as grounds for annulment.

Aftermath:

• The EU opened a formal investigation into TikTok under the Digital Services Act for potential failure to mitigate election manipulation risks.
• Romania held new elections in May 2025, with enhanced social media monitoring.
• The case became a precedent — the first documented instance where AI-enhanced interference triggered institutional response (election annulment), not just post-hoc analysis.

Key lessons:

• Institutional response is possible: Unlike most AI interference cases where impact is debated after the fact, Romania's institutions acted. This required classified intelligence evidence, however — suggesting that democratic response depends on intelligence capabilities most countries lack.
• TikTok's algorithm was the amplifier: The AI component was less about generating content and more about algorithmic amplification of coordinated accounts. Platform design can be as important as content generation.
• Low-profile elections are most vulnerable: Georgescu was a marginal candidate in a multi-candidate first-round election — exactly the scenario where small interventions can be decisive.
• Cryptocurrency complicates attribution: Foreign funding through crypto makes traditional campaign finance enforcement much harder.

Estimated impact: Decisive — a candidate polling at <5% won the first round, an outcome intelligence services attributed to the interference campaign.

Slovakia 2023: Deepfake Audio Incident

Event: Audio deepfake of liberal party leader discussing vote rigging surfaced days before election Result: Liberal party suffered upset loss Attribution: Unclear if deepfake was decisive

Analysis:

• Timing (just before election) maximized impact, minimized correction time
• Topic (vote rigging) highly salient and credible to some voters
• Close race amplified marginal effects

Estimated Impact: Possibly 1-3% vote shift, potentially decisive in close race

Lessons:

• Timing matters enormously
• Topic credibility affects impact
• Close races vulnerable to small effects

Taiwan 2024: Documented AI Influence Campaign

Event: Microsoft documented China-based AI-generated deepfakes targeting Taiwan election Result: Unclear impact on outcome Characteristics: First confirmed state-actor use of AI in foreign election

Analysis:

• Detected and publicized before election (reduced impact)
• Taiwan electorate somewhat prepared for Chinese interference
• Content quality varied (some obvious, some convincing)

Estimated Impact: <1% vote shift, not decisive

Lessons:

• Attribution and publicity can reduce impact
• Prepared electorates more resilient

Empirical Evidence Summary

The following table synthesizes experimental research on AI persuasion effects relevant to electoral contexts.

These findings suggest a paradox: while AI can produce highly persuasive content in experimental settings, real-world electoral effects remain difficult to detect. Possible explanations include: (1) experimental conditions differ from actual campaign contexts; (2) effects are real but small and distributed across many elections; (3) countervailing forces (skepticism, platform moderation) offset AI advantages in practice.

Quantitative Impact Estimates

Model 1: Multiplicative Probability

P(AI flips election) = P(close race) × P(AI campaign) × P(reaches voters) × P(shifts votes) × P(shift is decisive)

Where:
P(close race) = 0.15-0.30 (varies by election type)
P(AI campaign) = 0.50-0.90 (becoming common)
P(reaches voters) = 0.20-0.50 (platform moderation, virality)
P(shifts votes) = 0.05-0.15 (small persuasion effect)
P(shift is decisive) = 0.10-0.30 (in close race context)

Result: P(AI flips election) = 0.0015 to 0.054 (0.15% to 5.4%)

Interpretation: In any given election, AI disinformation has a ~0.2-5% chance of being decisive.

Over many elections (50+ major races in a year), AI disinformation likely flips 1-3 elections annually (current state).

Confidence: Very low. Enormous uncertainty in each parameter.

Model 2: Vote Margin Approach

Baseline Assumptions:

• 100 million voters
• 50-50 race
• 30% exposed to AI disinformation
• 5% of exposed shift votes
• 1.5 million vote shift (1.5% of total)

In close elections (decided by <1%): AI disinformation likely decisive

In moderate elections (3-5% margin): AI disinformation possibly influential but not clearly decisive

In landslide elections (>7% margin): AI disinformation unlikely decisive

Implication: ~20-30% of elections are close enough that AI disinformation could plausibly be decisive.

Scenario Analysis

The following scenarios represent distinct trajectories for AI disinformation's electoral impact over the 2025-2030 period.

The most concerning finding from recent research is the Romania 2024 case, where election results were annulled after evidence of AI-powered interference using manipulated videos. This represents the first documented case of AI disinformation being consequential enough to trigger institutional response.

flowchart TD
  subgraph Inputs["Model Inputs"]
      AI_CAP[AI Capability Growth<br/>30%/year quality improvement]
      PLAT[Platform Response<br/>Detection rate: 30-60%]
      REG[Regulatory Environment<br/>20 states with laws by 2024]
      VOTER[Voter Adaptation<br/>84% concerned about AI fakes]
  end

  subgraph Process["Impact Pathway"]
      GEN[Content Generation<br/>100-1000x cost reduction]
      DIST[Distribution<br/>1.5-4x reach vs traditional]
      PERS[Persuasion<br/>1.3-2x with personalization]
      VOTE[Vote Change<br/>5-15% of exposed shift]
  end

  subgraph Outcomes["Outcome Space"]
      IND[Individual Election<br/>0.2-5% flip probability]
      SYS[Systemic Trust<br/>2-5% annual decline]
      CRISIS[Democratic Crisis<br/>Conditional on accumulation]
  end

  AI_CAP --> GEN
  GEN --> DIST
  PLAT --> DIST
  DIST --> PERS
  PERS --> VOTE
  REG --> VOTE
  VOTER --> PERS
  VOTE --> IND
  VOTE --> SYS
  SYS --> CRISIS
  IND --> CRISIS

  style CRISIS fill:#ff6666
  style SYS fill:#ffaa66
  style IND fill:#ffdd66

Factors Moderating Impact

Increasing AI Impact

1. Targeting Sophistication: Better micro-targeting increases efficiency
2. Multimodal Content: Video/audio more persuasive than text
3. Coordination: Multiple AI campaigns from different sources reinforce messaging
4. Erosion of Trust: As authentic media becomes suspect, all information becomes equally (un)reliable
5. Authoritarian Backing: State-sponsored campaigns have more resources and persistence

Decreasing AI Impact

1. Platform Countermeasures: Detection, labeling, removal
2. Media Literacy: Educated populations more skeptical
3. Provenance Systems: C2PA and similar make authentic content verifiable
4. Partisan Polarization: Voters so entrenched that persuasion is difficult
5. Saturation: So much disinformation that all becomes noise

Trajectory Projections

2024-2026: Early Impact Phase

Characteristics:

• AI disinformation common but detectable
• Platforms implementing countermeasures
• Electorate beginning to adapt
• Estimated impact: 1-3% of close elections flipped

2026-2028: Escalation Phase

Characteristics:

• AI-generated content becomes harder to detect
• Personalization improves (better targeting)
• More actors deploy AI campaigns
• Public awareness increases but so does volume
• Estimated impact: 3-8% of close elections flipped

2028-2030: Saturation or Adaptation

Two Possible Paths:

Path A: Saturation (40% probability)

• So much disinformation that voters tune out
• All information treated as equally suspect
• Impact paradoxically decreases as volume increases
• Estimated impact: 2-5% of elections (impact declines)

Path B: Sophistication Wins (60% probability)

• Personalized, multimodal AI content highly effective
• Detection fails to keep pace
• Provenance systems not widely adopted
• Estimated impact: 10-20% of close elections flipped

Systemic Democratic Effects

Beyond individual elections, AI disinformation affects democratic health:

Trust Erosion:

• Even if specific election impacts are small, aggregate trust in media declines
• "Liar's dividend" makes all evidence deniable
• Democratic deliberation requires shared reality—this breaks down

Measured Impact:

• Trust in media: Declining 3-5% annually (accelerating)
• Belief in election integrity: Declining 2-4% annually
• Political polarization: Increasing (AI contribution unclear but likely 10-30%)

These systemic effects may matter more than vote margins in individual elections.

Policy Implications

If Impact is Currently Low (<2% of elections)

Interpretation: Current countermeasures working; worry may be overblown

Recommended Actions:

• Maintain current platform policies
• Monitor for increasing impact
• Continue media literacy efforts
• Avoid over-regulation that might harm free speech

If Impact is Moderate (2-8% of elections)

Interpretation: Significant threat but manageable with effort

Recommended Actions:

• Strengthen platform detection and removal
• Mandate provenance systems (C2PA)
• Increase funding for election security
• International cooperation on attribution and consequences

If Impact is High (>10% of elections)

Interpretation: Crisis-level threat to democratic integrity

Recommended Actions:

• Emergency measures: possible temporary restrictions on AI-generated political content
• Mandatory authentication for all political advertising
• Dramatic increase in election security budgets
• Consider election reforms (longer voting periods to allow fact-checking)

Beyond Influence: AI Threats to Election Infrastructure

The model above focuses on AI disinformation's influence on voter beliefs and choices. But elections can also be "stolen" through attacks on election infrastructure — the systems that register voters, count ballots, and report results. AI enhances these attack vectors in ways that the influence-focused model doesn't capture.

AI-Enhanced Cyberattacks on Election Systems

Election infrastructure presents a large attack surface: voter registration databases, electronic poll books, ballot marking devices, optical scanners, election management systems (EMS), and election night reporting (ENR) systems. AI enhances existing cyber threats in several ways:

Key insight: The most impactful infrastructure attacks may not change vote counts directly but rather create chaos and undermine confidence in results. An AI-generated deepfake of an election official announcing "irregularities" during vote counting could trigger a legitimacy crisis regardless of whether actual fraud occurred.

CISA assessment: The Cybersecurity and Infrastructure Security Agency has identified election infrastructure as critical national infrastructure since 2017. Their election security work has hardened many systems, but the attack surface remains large — over 8,000 election jurisdictions in the US alone, many with limited cybersecurity budgets and staffing.

Historical precedent: Russia's GRU targeted voter registration systems in all 50 US states in 2016 (per the Senate Intelligence Committee report), successfully breaching systems in at least two Florida counties. While no votes were changed, the intrusions demonstrated that state actors can access election infrastructure. AI could make such intrusions faster, harder to detect, and more precisely targeted.

AI-Powered Voter Roll Manipulation

Voter roll accuracy is a legitimate concern — states must maintain accurate lists under the National Voter Registration Act (NVRA). But AI dramatically changes the scale and precision of potential manipulation:

Automated mass voter challenges: AI can scan voter rolls and cross-reference them against commercial databases to generate mass challenges to voter eligibility. In Georgia, a single activist group challenged 364,000 voter registrations before the 2021 runoff elections. AI could automate and scale this process dramatically — generating thousands of challenges per day with superficially plausible justifications (address mismatches, name variations, etc.).

False-positive citizenship matching: The national citizenship database being built by DHS (linking voter rolls with immigration, SSA, and driver's license data) creates risks of algorithmic disenfranchisement. Name-matching algorithms have well-documented error rates, particularly for:

• Hispanic surnames (patronymic naming patterns create false matches)
• Asian names (transliteration variations)
• Hyphenated and compound names
• Common names in minority communities

Florida's 2000 voter purge (contracted to Database Technologies/DBT) erroneously flagged thousands of eligible voters as felons, disproportionately affecting Black voters. AI-powered matching at national scale could produce errors orders of magnitude larger.

Interstate Crosscheck problems as precedent: The Interstate Voter Registration Crosscheck Program, which compared voter rolls across 27 participating states, produced a 99% false-positive rate according to research by Stanford, Harvard, and Microsoft. The program matched voters using only first name, last name, and date of birth — flagging people like "James Brown" or "Maria Garcia" across multiple states. AI-powered systems could be more sophisticated but would still face fundamental statistical challenges when applied to 150+ million registered voters.

Probability estimate: AI-powered voter roll manipulation affecting election outcomes: 5-15% probability in any given US election cycle (higher in states with aggressive purge practices and limited judicial oversight).

AI-Powered Voter Suppression

Beyond disinformation that changes minds, AI enables targeted demobilization — discouraging specific groups from voting at all. This is distinct from the influence model above because the goal is not to persuade but to suppress turnout.

Micro-targeted demobilization: AI can identify persuadable non-voters in specific demographics and generate personalized content designed to depress turnout. Techniques include:

• Messages emphasizing futility ("your vote doesn't matter in this district")
• Targeted cynicism about specific candidates among their base
• Confusion about voting logistics (wrong dates, locations, ID requirements)
• AI-generated fake "official" communications about polling changes

The 2016 Cambridge Analytica operation demonstrated targeted demobilization using Facebook data — what they internally called "deterrence" campaigns aimed at African-American voters. AI would make such operations cheaper, faster, and harder to trace.

AI-generated robocalls: The January 2024 New Hampshire incident — where AI-cloned Biden voice urged 25,000 Democrats not to vote in the primary — demonstrated the basic capability. The FCC subsequently ruled AI-generated voices in robocalls illegal under the TCPA in February 2024, and the perpetrator (Steve Kramer, a political consultant) was fined $6 million by the FCC and faced criminal charges in New Hampshire. But enforcement is reactive — the calls went out before anyone could stop them.

Automated poll worker harassment: AI could automate campaigns to intimidate poll workers through personalized threatening messages, deepfake "evidence" of misconduct, or coordinated social media harassment. A 2024 Brennan Center survey found more than 1 in 3 election officials have faced harassment, abuse, or threats, with 1 in 4 worried about being assaulted at home or work. AI-powered harassment could accelerate the resulting attrition of experienced election staff.

Quantitative estimate: A 2026 PNAS study tracked 10,000+ participants who installed an app capturing every ad viewed for six weeks before the 2016 election. Key findings:

• Participants who saw vote-suppressing Facebook messages were 1.9% less likely to actually vote
• Nonwhite voters in battleground state minority counties were 14.2% less likely to vote compared to white voters in non-battleground areas
• Extrapolated nationally, approximately 4.7 million people may have been kept from voting
• This was from crude, human-written 2016-era content — AI personalization would likely produce larger effects

In close elections, even 1-2% turnout reductions in targeted demographics are decisive. Black and Latino voters, young voters, and voters in precincts with fewer polling locations are most vulnerable.

Combined Threat Assessment

The influence model (previous sections) and infrastructure threats (this section) are not independent — they can compound:

A sophisticated actor — particularly a state actor with intelligence capabilities — could deploy all three simultaneously: influence operations to shift preferences, infrastructure attacks to create chaos and undermine confidence, and suppression campaigns to reduce opposition turnout. The combined effect would be substantially larger than any single vector.

Model Limitations

This model faces fundamental measurement challenges that limit confidence in its estimates.

Counterfactual Problem. The core limitation is that we cannot observe what would have happened without AI disinformation in any given election. Romania 2024 provides suggestive evidence, but even there, the annulment was based on evidence of interference, not proof of decisive impact. Every estimate in this model involves a counterfactual comparison that cannot be directly observed.

Multi-Causality and Attribution. Elections are influenced by dozens of factors: economic conditions, candidate quality, campaign spending, media coverage, debates, and ground operations. Isolating the marginal contribution of AI disinformation from this complex system is methodologically challenging. The meta-analysis of 49 field experiments finding zero average effect from campaign contact illustrates how difficult persuasion measurement is even for well-controlled interventions.

Detection Bias. We can only measure detected AI campaigns. The most sophisticated operations may go entirely unnoticed, meaning our estimates potentially undercount the most impactful instances. Conversely, the Knight Columbia analysis of 78 election deepfakes found that 39 had no deceptive intent, suggesting overcount in some datasets.

Heterogeneity. Impact varies dramatically by context: election type (presidential vs. local), electorate characteristics (polarization level, media literacy), and institutional environment (platform policies, legal frameworks). Parameter estimates that work for U.S. presidential elections may be inappropriate for local ballot initiatives or elections in developing democracies.

Rapid Technological Change. Both AI generation capabilities and detection methods are improving rapidly. Model parameters derived from 2024 data may be obsolete by 2026. The finding that "cheap fakes" outperformed AI 7:1 in 2024 may not hold as AI quality improves and costs fall further.

Key Debates

Did AI "Break" 2024 Elections? Research suggests no, but measurement problems make this uncertain. Absence of evidence is not evidence of absence.

What Matters More: Individual Elections or Systemic Trust? Even if AI doesn't flip many elections, erosion of epistemic commons might be the bigger harm.

Can Democracy Survive in an Era of Undetectable Disinformation? Pessimists say no; optimists argue humans have adapted to information threats before.

Related Models

• Disinformation Detection Arms RaceAnalysisDisinformation Detection Arms Race ModelModels adversarial dynamics between AI generation and detection of synthetic content, projecting detection accuracy will fall from 65% (2024) to ~50% (near-random) by 2030 under medium adversarial ...Quality: 55/100 - Can we detect it at all?
• Deepfakes Authentication CrisisAnalysisDeepfakes Authentication Crisis ModelProjects authentication crisis threshold when detection accuracy falls to 50-55% (chance levels): audio by 2026-2027, images 2025-2027, video 2026-2030. Content provenance (C2PA) has 30% probabilit...Quality: 53/100 - Visual media authenticity

Sources

AI Disinformation Research

• Goldstein, J. et al. "How persuasive is AI-generated propaganda?" PNAS Nexus (2024). Found GPT-3 can create propaganda as persuasive as human-written content with minimal effort.
• Matz, S. et al. "The potential of generative AI for personalized persuasion at scale." Scientific Reports (2024). Demonstrated 1.3-2x persuasion uplift from AI personalization across 7 studies (N=1,788).
• Bai, H. et al. "Persuading voters using human-artificial intelligence dialogues." Nature (2025). Pre-registered experiments showing AI dialogues produce larger effects than traditional video ads.

Electoral Impact Studies

• Kalla, J. & Broockman, D. "The Minimal Persuasive Effects of Campaign Contact in General Elections." American Political Science Review (2018). Meta-analysis of 49 field experiments finding ~zero average effect.
• Simon, F. & Camargo, C. "We Looked at 78 Election Deepfakes." Knight Columbia (2024). Found cheap fakes 7x more common than AI deepfakes; 39 of 78 cases had no deceptive intent.
• CIGI. "Then and Now: How Does AI Electoral Interference Compare in 2025?" Comprehensive comparison including Romania 2024 annulment case.

Voter Suppression Research

• "Targeted digital voter suppression efforts likely decrease voter turnout." PNAS (2026). 10,000+ participants tracked over 6 weeks; found 1.9% turnout reduction from suppression ads, 14.2% for nonwhite voters in battleground minority counties. Estimated ~4.7 million kept from voting in 2016.
• Brennan Center for Justice. "Preparing to Fight AI-Backed Voter Suppression." Analysis of AI-powered suppression tactics and countermeasures.
• ProPublica. "Six Right-Wing Activists Filed 89,000 Georgia Voter Roll Challenges." (2023). Documenting approximately 100,000 automated voter challenges in Georgia by a small network of activists.
• NPR. "Voter registration mass challenges fueled by EagleAI." (2024). Documenting EagleAI-driven voter challenges in Georgia counties.

Platform and Social Media Effects

• Aral, S. & Eckles, D. "Protecting elections from social media manipulation." Science (2019). Proposed research agenda for measuring manipulation effects.
• Allcott, H. et al. "The effects of Facebook and Instagram on the 2020 election." PMC (2024). Deactivation experiment (N=35,000) finding limited effect on political views.
• Harvard Kennedy School Misinformation Review (2024). Analysis of why predicted AI impacts in 2024 did not materialize.