Public Opinion

{
  "id": "public-opinion",
  "numericId": null,
  "path": "/knowledge-base/metrics/public-opinion/",
  "filePath": "knowledge-base/metrics/public-opinion.mdx",
  "title": "Public Opinion & Awareness",
  "quality": 52,
  "importance": 52,
  "contentFormat": "article",
  "tractability": null,
  "neglectedness": null,
  "uncertainty": null,
  "causalLevel": null,
  "lastUpdated": "2026-01-28",
  "llmSummary": "Comprehensive survey compilation showing AI concern rising rapidly (37%→50%, 2021-2025) with strong regulatory support (70-80%) but massive literacy gap (99% use AI, 39% aware). Only 12% mention existential risk unprompted despite 69% supporting development pause, suggesting concern focuses on near-term harms; trust declining across institutions with 45-point gap between China (72%) and US (32%).",
  "structuredSummary": null,
  "description": "Tracking public understanding, concern, and attitudes toward AI risk and safety",
  "ratings": {
    "novelty": 3.5,
    "rigor": 6,
    "actionability": 4.5,
    "completeness": 7
  },
  "category": "metrics",
  "subcategory": null,
  "clusters": [
    "ai-safety",
    "epistemics"
  ],
  "metrics": {
    "wordCount": 2566,
    "tableCount": 4,
    "diagramCount": 0,
    "internalLinks": 21,
    "externalLinks": 2,
    "footnoteCount": 0,
    "bulletRatio": 0.63,
    "sectionCount": 74,
    "hasOverview": true,
    "structuralScore": 9
  },
  "suggestedQuality": 60,
  "updateFrequency": 45,
  "evergreen": true,
  "wordCount": 2566,
  "unconvertedLinks": [],
  "unconvertedLinkCount": 0,
  "convertedLinkCount": 10,
  "backlinkCount": 2,
  "redundancy": {
    "maxSimilarity": 16,
    "similarPages": [
      {
        "id": "structural",
        "title": "Meta & Structural Indicators",
        "path": "/knowledge-base/metrics/structural/",
        "similarity": 16
      },
      {
        "id": "economic-labor",
        "title": "Economic & Labor Metrics",
        "path": "/knowledge-base/metrics/economic-labor/",
        "similarity": 14
      },
      {
        "id": "expert-opinion",
        "title": "Expert Opinion",
        "path": "/knowledge-base/metrics/expert-opinion/",
        "similarity": 14
      },
      {
        "id": "ai-impacts",
        "title": "AI Impacts",
        "path": "/knowledge-base/organizations/ai-impacts/",
        "similarity": 14
      },
      {
        "id": "trust-erosion-dynamics",
        "title": "Trust Erosion Dynamics Model",
        "path": "/knowledge-base/models/trust-erosion-dynamics/",
        "similarity": 13
      }
    ]
  }
}

{
  "id": "public-opinion",
  "type": "ai-transition-model-metric",
  "title": "Public Opinion",
  "description": "Metrics tracking public awareness, concern levels, and trust regarding AI systems and AI safety.",
  "tags": [
    "public",
    "surveys",
    "trust"
  ],
  "relatedEntries": [
    {
      "id": "societal-trust",
      "type": "ai-transition-model-parameter",
      "relationship": "measures"
    },
    {
      "id": "preference-authenticity",
      "type": "ai-transition-model-parameter",
      "relationship": "measures"
    }
  ],
  "sources": [],
  "lastUpdated": "2025-12",
  "customFields": []
}

No facts for this entity

No external links

{
  "title": "Public Opinion & Awareness",
  "description": "Tracking public understanding, concern, and attitudes toward AI risk and safety",
  "sidebar": {
    "order": 8
  },
  "importance": 52,
  "update_frequency": 45,
  "lastEdited": "2026-01-28",
  "quality": 52,
  "llmSummary": "Comprehensive survey compilation showing AI concern rising rapidly (37%→50%, 2021-2025) with strong regulatory support (70-80%) but massive literacy gap (99% use AI, 39% aware). Only 12% mention existential risk unprompted despite 69% supporting development pause, suggesting concern focuses on near-term harms; trust declining across institutions with 45-point gap between China (72%) and US (32%).",
  "ratings": {
    "novelty": 3.5,
    "rigor": 6,
    "actionability": 4.5,
    "completeness": 7
  },
  "clusters": [
    "ai-safety",
    "epistemics"
  ]
}

---
title: "Public Opinion & Awareness"
description: "Tracking public understanding, concern, and attitudes toward AI risk and safety"
sidebar:
  order: 8
importance: 52
update_frequency: 45
lastEdited: "2026-01-28"
quality: 52
llmSummary: "Comprehensive survey compilation showing AI concern rising rapidly (37%→50%, 2021-2025) with strong regulatory support (70-80%) but massive literacy gap (99% use AI, 39% aware). Only 12% mention existential risk unprompted despite 69% supporting development pause, suggesting concern focuses on near-term harms; trust declining across institutions with 45-point gap between China (72%) and US (32%)."
ratings:
  novelty: 3.5
  rigor: 6
  actionability: 4.5
  completeness: 7
clusters: ["ai-safety", "epistemics"]
---
import {R, EntityLink, DataExternalLinks} from '@components/wiki';



## Key Links

| Source | Link |
|--------|------|
| Official Website | [wikiedu.org](https://wikiedu.org/blog/2025/02/06/how-i-learned-to-stop-worrying-and-love-wikipedia/) |
| Wikipedia | [en.wikipedia.org](https://en.wikipedia.org/wiki/Public_opinion) |

<DataExternalLinks pageId="public-opinion" />

## Overview

Public opinion shapes the political feasibility of <EntityLink id="E608">AI governance</EntityLink> and influences the trajectory of AI development through democratic pressure, consumer behavior, and social norms. This page tracks key metrics on public awareness, concern, trust, and literacy regarding AI risks and capabilities.

**Key Finding**: While AI awareness is nearly universal (95%+ in the US), **specific awareness of existential risk remains low** (~12% mention it unprompted), though **general concern is rising rapidly** (from 37% in 2021 to 50% in 2025).

---

## 1. Population Aware of AI Existential Risk

### Specific X-Risk Awareness
- **12.3%** of US adults mention AI as a potential cause of human extinction (unprompted)
  - Source: <EntityLink id="E558">Rethink Priorities</EntityLink> survey, April 2023 (n=300)
  - **+78% increase** from 6.9% in December 2022
  - Growth attributed to ChatGPT media coverage surge

- **43%** are very/somewhat concerned about AI causing the end of humanity
  - Source: YouGov, June 2025
  - **16%** very concerned, **27%** somewhat concerned
  - Up from **37%** in March 2025

- **59%** of US adults support prioritizing mitigating extinction risk from AI
  - Source: Rethink Priorities online poll, 2023
  - **26%** disagree
  - Among those who disagree: 36% cite "other priorities," 23% say "not extinction," 18% say "not yet"

### Ranking Among Existential Threats
- **4%** select AI as the most likely cause of human extinction
  - Ranks below nuclear war (42%), climate change, pandemics, and asteroids
  - Source: Rethink Priorities nationally-representative survey

### Expert vs. Public Gap
- **AI researchers** estimate median **5%** chance of human extinction from AI
  - Source: <EntityLink id="E512">AI Impacts</EntityLink> survey of 2,778 researchers, 2023
  - ~**40%** of researchers indicate >10% chance of catastrophic outcomes
  - Unchanged from 2022 survey (same 5% median)

---

## 2. Population Concerned About AI Safety (General)

### Overall Concern Trends (US)
- **50%** of Americans are more concerned than excited about AI (2025)
  - Source: Pew Research, survey of 5,023 adults, June 9-15, 2025
  - **+35% increase** from 37% in 2021
  - Only **11%** are more excited than concerned

- **57%** rate societal risks of AI as high
  - Source: Pew Research, August 2024 (n=5,410)
  - vs. **25%** who rate benefits as high

- **47%** believe AI effects on society will be negative
  - Source: YouGov, June 2025
  - **+38% increase** from 34% in December 2024
  - Steady upward trend: 34% → 40% → 41% → 47% over 6 months

### Trust in AI Systems
- **31%** of Americans trust AI (2% fully, 29% somewhat)
  - Source: Gallup, 2025 (n=3,128)
  - **60%** do not trust AI to make fair, unbiased decisions
  - Only **8%** consider themselves "very knowledgeable" about AI

- **Trust is declining**: 25% say trust decreased in past year
  - vs. 21% increased, 47% unchanged
  - Source: YouGov, 2025

- **18%** would trust AI to make decisions/take actions
  - vs. **53%** would not
  - Even among Gen Z: 43% mistrustful vs. 26% trustful

### Specific Concerns (US Public vs. AI Experts)
| Concern | Public | AI Experts |
|---------|--------|------------|
| Less human connection | 57% | 37% |
| Job loss | 56% | 25% |
| More concerned than excited | 51% | 15% |
| Negative impact next 20 years | 83% | 44% |

*Source: Pew Research, August 2024*

### Global Variation in Concern
**Most concerned countries** (>50% more concerned than excited):
- United States
- Italy
- Australia
- Brazil
- Greece

**Least concerned**:
- South Korea (16%)
- China (72% trust AI - highest globally)

*Source: Pew Research Global Survey, October 2025*

---

## 3. AI Risk Media Coverage

### Article Volume
- **38,787 articles** on AI from 12 major English newspapers (2010-2023)
  - Source: Ittefaq et al. (2024), analysis of 12-country media coverage
  - Dramatic increase in recent years, especially post-ChatGPT

- **24,827+ articles** on generative AI specifically (Jan 2018 - Nov 2023)
  - Source: Landscape of Generative AI in Global News (2024)
  - Sharp spike after ChatGPT launch (November 2022)

### Academic Publications on AI in Journalism
- **2023**: 39 publications
- **2024**: 106 publications (**+172% year-over-year**)
- **2020-2022**: ~20-25 publications/year (plateau)

*Source: Systematic bibliometric analysis (2016-2024)*

### Media Sentiment
- **21%** negative coverage
- **13%** positive coverage
- **66%** neutral coverage

*Source: 12-country analysis (2010-2023)*

### Trend in Critical Coverage
- UK and US media have become **more critical over time**
- Progressive outlets (NYT, The Guardian) "going slightly negative about AI each year"
- Increased references to risks and concerns

---

## 4. AI Safety Google Search Trends

### General AI Search Interest
- **+250% year-over-year** increase in "artificial intelligence" searches (UK, 2023)
  - Source: Think with Google, 2023

- **Peak interest**: October 2023 for "AI-powered search"
  - Stabilized in first half of 2024
  - Source: Statista, Google Trends data

### Generative AI Specific
- **Peak**: Week ending March 3, 2024 (score: 100)
  - Surge from mid-February to early March 2024
  - Source: Statista, Google Trends

### Search Behavior Shift
- Searches evolving from **curiosity** → **practical application**
- "How to AI" queries growing **faster** than basic informational queries
- **Hands-on queries** dominating over definitional searches

### 2025 Top Trending AI Searches
- **#1 globally**: "Gemini" (Google AI assistant)
- AI-generated content: Barbie AI, action figures, Ghibli-style art
- Indicates mainstream adoption of AI tools

*Source: Google Year in Search 2025*

---

## 5. Trust in AI Companies

### Overall Company Trust
- **79%** of Americans don't trust companies to use AI responsibly
  - Source: Bentley-Gallup Survey, 2025

- **47%** globally confident AI companies protect personal data (2024)
  - **Down from 50%** in 2023 (**-6%**)
  - Source: Ipsos, cited in Stanford AI Index 2025

### Trust by Institution Type (US)
| Institution | Trust Level |
|-------------|-------------|
| Employers | 71% |
| Universities | 67% |
| Large tech companies | 61% |
| Start-ups | 51% |

*Source: McKinsey US employee survey, Oct-Nov 2024*

### Global Trust Trends
- **Global average**: Only **46%** willing to trust AI systems
  - Source: KPMG Global AI Trust Study (n=48,000 across 47 countries), 2025

- **Advanced economies**: Trust drops to **39%**
  - Trust declining in wealthy nations, rising in emerging economies

### Trust Metrics Declining (2022-2024)
- Perceived trustworthiness: **63% → 56%** (**-11%**)
- Willingness to rely on AI: **52% → 43%** (**-17%**)
- Worried about AI: **49% → 62%** (**+27%**)

*Source: Global trust surveys (2022-2024)*

### Regional Extremes
- **Highest trust**: China (72%), India, Nigeria
- **Lowest trust**: US (32%), Australia, Ireland, Netherlands (under 33%)

*Source: Edelman Trust Barometer 2025*

### Experience Effect
- **AI users**: 46% trust AI
- **Non-users**: 23% trust AI
- **Trust doubles with usage**

*Source: Gallup, 2025*

---

## 6. Trust in AI Regulation

### Government Regulatory Trust
- **62%** of US public have little/no confidence in government to regulate AI
  - Source: Pew Research, August 2024 (n=5,410)
  - **53%** of AI experts also lack confidence

### Demand for Regulation
- **80%** say government should maintain safety rules even if AI develops more slowly
  - Source: Gallup/SCSP survey, 2025 (n=3,128)
  - Only **9%** prioritize speed over safety
  - Bipartisan support: 88% Democrats, 79% Republicans/Independents

- **71%** believe regulation is needed (41% "much more", 30% "somewhat more")
  - Source: YouGov, 2025
  - **Up from 64%** in December 2024 (**+11%**)

- **70%** globally believe national/international AI regulation is needed
  - Source: KPMG Global AI Trust Study, 2025

### Perceived Regulatory Inadequacy
- **29%** of US consumers believe current regulations are sufficient
  - **72%** say more regulation needed
  - **81%** would trust AI more if laws/policies were in place

### Government Role Support
- **~50%** of Americans agree government should have major role in AI oversight (end of 2024)
  - Source: Ipsos/Stanford AI Index, 2024

---

## 7. Support for AI Development Pause

### Public Support for 6-Month Pause (2023)
- **69%** support a 6-month pause on some AI development
  - **41%** strongly support
  - **28%** somewhat support
  - **13%** oppose (4% strongly, 9% somewhat)
  - **18%** not sure

*Source: YouGov poll, April 3, 2023 (n=20,810 US adults)*

### Support for Government-Enforced Moratorium
- **Average 73%** support (Yes + Maybe)
  - **39%** definite "Yes" (average)
  - Peak support after media intervention: **54%** (CNBC survey), **54%** (CNN survey)

- **Government enforcement**: Average **69%** support (Yes + Maybe)
  - **35%** definite "Yes"
  - Peak: **52%** (CNBC), **44%** (CNN)

*Source: EA Forum study, April 2023 (n=300, multiple survey conditions)*

### Context: <EntityLink id="E528">Future of Life Institute</EntityLink> Open Letter
- **March 2023**: Open letter calling for 6-month pause on systems >GPT-4
- **30,000+ signatures** including <EntityLink id="E380">Yoshua Bengio</EntityLink>, <EntityLink id="E290">Stuart Russell</EntityLink>, <EntityLink id="E116">Elon Musk</EntityLink>
- **Impact**: Generated "renewed urgency within governments," normalized expressing AI fears
- **Reality**: No pause occurred; investments in large models continued

---

## 8. AI Literacy Rate by Demographic

### Self-Reported Understanding
- **67%** globally say they have "good understanding" of AI (2024)
  - Source: Ipsos AI Monitor, 32-country survey

### By Generation (Global)
| Generation | Good Understanding |
|------------|-------------------|
| Gen Z | 72% |
| Millennials | 71% |
| Gen X | ≈60-65% (est.) |
| Baby Boomers | 58% |

*Source: Ipsos AI Monitor 2024*

### Actual vs. Perceived Knowledge (US)
- **98%** have heard about AI
- **39%** report using AI
- **BUT**: When asked about 6 common AI products, **99%** have used at least one
  - **83%** used 4+ AI products
  - Indicates **severe awareness gap** about what counts as AI

*Source: Gallup, 2025*

### Depth of Understanding (US)
- **34.13%** highly familiar with AI
- **50.08%** somewhat familiar
- **15.79%** know nothing about AI

- **Among those "familiar"**:
  - Only **13.73%** actually understand AI processes
  - **57.8%** have some understanding
  - **28.47%** just know the term

*Source: Survey of 800 Americans, 2024*

### University Students (4 Asian/African nations)
- **Average AI literacy**: 2.98 out of 5 (moderate skill level)
- **Significant disparities** by:
  - Nationality
  - Field of study (technical > non-technical)
  - Academic degree level
- **No significant difference** by:
  - Gender
  - Age

*Source: Comparative transnational survey (n=1,800), 2024*

### Gender Gap in Generative AI Use
- **Persistent global gender gap** in GenAI usage
- **Male students** show more trust in AI than female students (UK & Poland study, 2024)
- **Younger adults** (under 45) more open: 36% expect positive impact vs. 19% of 45+

### Urban-Rural Divide
- **Rural populations** perceive higher privacy/safety risks
- **Urban populations** more accepting of AI
- Functional reliability concerns differ significantly

### Student Preparedness
- **58%** of students don't feel they have sufficient AI knowledge/skills
- **48%** don't feel prepared for AI-enabled workplace
- Despite high usage rates

*Source: DEC Global AI Student Survey 2024*

---

## 9. Accuracy of Public Beliefs About AI Capabilities

### Common Misconceptions
**Top misconceptions identified** (Survey of 800 Americans, 2024):
1. "AI is fully autonomous and self-learning" (**50%** need clarification)
2. "AI makes decisions without any errors" (**50%** need clarification)

### Overestimation vs. Underestimation
- **Public tends to both overestimate and underestimate** AI capabilities
- **Overestimate**: Autonomy, reasoning, consciousness
- **Underestimate**: Current practical capabilities, scope of existing AI use

### Understanding of AI Ubiquity
- **Awareness gap**: 39% report using AI, but 99% actually use AI-enabled products
- **61% unaware** they use AI regularly
- Common products not recognized as AI:
  - Navigation apps
  - Streaming recommendation engines
  - Social media algorithms
  - Weather forecasting
  - Online shopping personalization

*Source: Gallup, 2025*

### Expert vs. Public Perception Gap
- **Experts 3x more optimistic** about AI impact (56% vs. 17%)
- **Experts 4x more excited** than public (47% vs. 11%)
- **Largest perception gaps**:
  - Long-term societal impact
  - Job market effects
  - Loss of human connection

*Source: Pew Research, August 2024*

### Misinformation Concerns
- **83.4%** of Americans concerned about AI spreading misinformation in 2024 election
  - Source: US public opinion survey, August 2023

### Desire for Better Information
**What Americans want to learn more about** (2024 survey):
1. **57.05%**: Accuracy of AI-generated results
2. **56.96%**: Data security when using AI
3. **48.22%**: How AI makes decisions

### Alignment of Beliefs with Reality
- **Limited research** specifically quantifying accuracy of public beliefs
- Most studies focus on **awareness** and **attitudes**, not **correctness**
- Significant need for systematic assessment of belief accuracy

---

## Key Trends & Insights

### 1. Rapid Concern Growth (2021-2025)
- Concern increased **35%** in 4 years (37% → 50%)
- Acceleration in 2024-2025: 34% → 47% in 6 months
- **ChatGPT** (Nov 2022) identified as major inflection point

### 2. High Pause Support but Low X-Risk Awareness
- **69%** support development pause
- Only **12%** mention existential risk unprompted
- Suggests concern is about **near-term harms**, not extinction

### 3. Erosion of Trust
- Trust in AI systems: **declining**
- Trust in companies: **declining** (50% → 47%)
- Trust in government to regulate: **low** (62% lack confidence)
- **But**: Experience builds trust (46% vs. 23%)

### 4. Massive Literacy Gap
- **99%** use AI products
- **39%** think they use AI
- **60-point awareness gap** about everyday AI

### 5. Expert-Public Divergence
- **3x** gap in optimism about long-term impact
- **4x** gap in excitement vs. concern
- Suggests communication challenge for AI safety advocates

### 6. Global Variation
- **Emerging markets**: High trust, high optimism (China 72% trust)
- **Advanced economies**: Low trust, high concern (US 32% trust)
- **45-point trust gap** between China and US

### 7. Strong Support for Regulation
- **70-80%** want government to prioritize safety over speed
- **Bipartisan consensus** (88% Dems, 79% Reps)
- But **low confidence** government can deliver (62% skeptical)

---

## Measurement Challenges

### 1. Question Framing Effects
- "AI" vs. "artificial intelligence" vs. "machine learning" elicits different responses
- "Existential risk" vs. "very bad outcomes" vs. "human extinction" varies widely
- Media exposure immediately before survey significantly affects responses

### 2. Awareness of AI Use
- People don't recognize AI in everyday products
- Self-reported usage dramatically underestimates actual usage
- Complicates measuring "literacy" vs. "awareness"

### 3. Temporal Volatility
- Opinions shift rapidly with news cycles
- ChatGPT caused immediate awareness spike
- Media interventions show immediate effect (EA Forum study)

### 4. Sample Representativeness
- Online panels vs. representative samples
- US-centric data (most surveys)
- Limited longitudinal tracking with consistent methodology

### 5. Correlation vs. Causation
- Does experience increase trust, or do trusting people seek experience?
- Does media coverage increase concern, or does concern drive coverage?
- Difficult to establish causal mechanisms

---

## Data Sources

### Primary Survey Organizations
- **<R id="40fcdcc3ffba5188">Pew Research Center</R>** - US public & AI experts (2024-2025)
- **<R id="f36d4b20ce95472c">YouGov</R>** - US tracking surveys (2024-2025)
- **<R id="24a4a75bbecb01ca">Gallup</R>** - Trust and awareness (2023-2025)
- **<R id="7f85b24f262562ec">Ipsos</R>** - Global AI Monitor (30+ countries, 2024-2025)
- **<R id="2f254d7fc3f63c7f">KPMG Global AI Trust Study</R>** - 47 countries, 48,000 respondents (2025)

### Academic & Nonprofit
- **<R id="6ff0c861116fa036">Rethink Priorities</R>** - AI policy & x-risk awareness
- **<R id="a0e5c1ff413bb7d8">AI Impacts</R>** - Expert surveys (2,778 researchers, 2023)
- **<R id="d2b4293d703f4451">Stanford HAI AI Index</R>** - Comprehensive annual report

### Media & Trends Analysis
- **<R id="25564b0c33b8d4bb">Google Trends</R>** - Search behavior
- **<R id="7799bbdc58fe571f">Ittefaq et al. (2024)</R>** - 12-country media analysis (38,787 articles)

---

## Related Metrics

- **<EntityLink id="E132">Expert Opinion</EntityLink>** - AI researcher surveys, P(doom) estimates
- **<EntityLink id="E154">Governance & Policy</EntityLink>** - Regulatory responses to public opinion
- **<EntityLink id="E184">Lab Behavior</EntityLink>** - How public pressure affects AI companies
- **<EntityLink id="E289">Structural Indicators</EntityLink>** - Information ecosystem quality

---

## Last Updated
December 24, 2025

**Note**: This page synthesizes data from multiple surveys conducted 2023-2025. Survey methodologies, sample sizes, and question wordings vary significantly. Numbers should be interpreted as indicative trends rather than precise measurements. For specific use cases, consult original sources for methodology details.