R0029/2026-03-27/Q002 — Assessment¶
BLUF¶
Public sentiment toward AI-generated content is deeply fragmented rather than uniformly positive or negative. Global trust stands at only 46% (KPMG, 48K respondents), but this masks a dramatic split: 39% in advanced economies versus 57% in emerging economies. A trust-use paradox exists where 66% of people use AI regularly despite majority distrust. Attitudes are trending slowly positive (52% to 55% seeing AI as beneficial, 2022-2024) but remain far from consensus.
Probability¶
Rating: N/A — this is a descriptive question, not a probability assessment
Confidence in assessment: High
Confidence rationale: Three independent large-scale surveys (KPMG/Melbourne 48K+, Stanford HAI aggregation, Ipsos 26-country longitudinal) converge on the same finding: fragmented, context-dependent sentiment. The convergence of independent data sources with large sample sizes provides strong confidence.
Reasoning Chain¶
- The KPMG/Melbourne study (48,000+ respondents, 47 countries) finds only 46% globally willing to trust AI, with a 18-point gap between advanced (39%) and emerging (57%) economies. [SRC01-E01, High reliability, High relevance]
- The Stanford HAI AI Index 2025 aggregates Ipsos data showing country-level ranges from 36% (Netherlands) to 83% (China) seeing AI as beneficial — a 47-point spread. [SRC02-E01, High reliability, High relevance]
- Ipsos longitudinal data shows a modest positive trend: 52% to 55% seeing AI as beneficial (2022-2024), with 18 of 26 countries trending upward. [SRC03-E01, High reliability, High relevance]
- JUDGMENT: The trust-use paradox (66% use, 46% trust) is the most analytically important finding. It suggests adoption is driven by perceived necessity or workplace pressure rather than positive sentiment, which has implications for how AI disclosure will be received.
- JUDGMENT: No single characterization (positive/negative) captures reality. The dominant pattern is fragmentation along geographic, economic, and contextual lines.
Evidence Base Summary¶
| Source | Description | Reliability | Relevance | Key Finding |
|---|---|---|---|---|
| SRC01 | KPMG/Melbourne global study | High | High | 46% trust; 66% use; 57% hide AI work |
| SRC02 | Stanford HAI AI Index 2025 | High | High | 36-83% range; slow positive trend |
| SRC03 | Ipsos 26-country longitudinal | High | High | 52% to 55% beneficial (2022-2024) |
Collection Synthesis¶
| Dimension | Assessment |
|---|---|
| Evidence quality | Robust — three large-scale, methodologically sound surveys with combined N > 100,000 |
| Source agreement | High — all sources converge on fragmented, context-dependent sentiment |
| Source independence | High — KPMG/Melbourne, Stanford HAI, and Ipsos are independent organizations |
| Outliers | None — all sources tell the same story |
Detail¶
The three major surveys are genuinely independent: KPMG/Melbourne conducted their own 48K-person survey, Stanford HAI aggregated multiple independent data sources, and Ipsos conducted longitudinal polling across 26 countries. All three converge on the same finding. The most important insight is not any single number but the fragmentation pattern itself — attitudes depend more on where you live and what context you're in than on any universal human reaction to AI.
Gaps¶
| Missing Evidence | Impact on Assessment |
|---|---|
| Content-type-specific attitudes (text vs. image vs. code) | Cannot distinguish whether AI-generated text is perceived differently from AI-generated images |
| Technology community vs. general public breakdown | The query asks about both but surveys mostly report general population data |
| Attitudes toward AI-generated content specifically (vs. AI generally) | Most surveys measure attitudes toward "AI" broadly, not "AI-generated content" specifically |
Researcher Bias Check¶
Declared biases: No researcher profile provided for this run.
Influence assessment: The query contains an embedded assumption — "including trust levels and negative attitudes" — which presupposes negative attitudes exist. This was surfaced and tested rather than assumed. The evidence confirmed that negative attitudes exist in specific contexts but are not universal.
Cross-References¶
| Entity | ID | File |
|---|---|---|
| Hypotheses | H1, H2, H3 | hypotheses/ |
| Sources | SRC01, SRC02, SRC03 | sources/ |
| ACH Matrix | — | ach-matrix.md |
| Self-Audit | — | self-audit.md |