Reduced AI Acceptance After the Generative AI Boom: Evidence From a Two-Wave Survey Study

TL;DR

This study examines how public acceptance of AI shifted in Switzerland after the GenAI boom catalyzed by ChatGPT. Using two representative survey waves, the authors show a general decline in AI acceptability and a rising preference for human oversight across multiple decision scenarios, with the strongest effects in high-stakes domains. The analysis also reveals amplification of pre-existing social inequalities, notably across education levels, language regions, and gender, particularly in health-related contexts. The findings suggest that AI governance and regulation should be context-sensitive and inclusive, acknowledging evolving public preferences and the risk of widening digital divides as GenAI technologies expand.

Abstract

The rapid adoption of generative artificial intelligence (GenAI) technologies has led many organizations to integrate AI into their products and services, often without considering user preferences. Yet, public attitudes toward AI use, especially in impactful decision-making scenarios, are underexplored. Using a large-scale two-wave survey study (n_wave1=1514, n_wave2=1488) representative of the Swiss population, we examine shifts in public attitudes toward AI before and after the launch of ChatGPT. We find that the GenAI boom is significantly associated with reduced public acceptance of AI (see Figure 1) and increased demand for human oversight in various decision-making contexts. The proportion of respondents finding AI "not acceptable at all" increased from 23% to 30%, while support for human-only decision-making rose from 18% to 26%. These shifts have amplified existing social inequalities in terms of widened educational, linguistic, and gender gaps post-boom. Our findings challenge industry assumptions about public readiness for AI deployment and highlight the critical importance of aligning technological development with evolving public preferences.

Figures (9)

• Figure 1: The generative AI boom in Swiss media and politics. The left panel shows the fraction of AI news in all Swiss news from 2000 to 2025 that mention AI (weighted by the total number of news per year). The left panel shows the number of parliamentary interventions on AI from 2015 to 2024 involving AI-related terms.
• Figure 2: The generative AI boom is associated with a significant increase in required human control for five out of seven scenarios. In both survey waves, human control preferences are heavily dependent on context. Just as in Figure \ref{['fig:average_acceptance_comparison_by_scenario']}, we show weighted means with 95% CI and significance levels indicate statistically significant differences across survey waves as determined by a two-sample design-based t-test.
• Figure 3: Response distribution in % for acceptance questions across survey waves and scenarios.
• Figure 4: Response distribution in % for human control questions across survey waves and scenarios. Notice that in this figure, the two answer options (AI decision with human oversight and AI-assisted human decision) are merged into one single category titled human-AI collaboration, as they are conceptually very similar -- see Table \ref{['tab:human_control_answers']} for more details. In Appendix \ref{['app:human_control_without_grouped_answer']}, we provide the same overview but consider both answer options separately, showing that the overall results remain unaffected. However, notice that the overall reduction of the share of people favoring human-AI collaboration is mainly driven by fewer respondents opting for AI decisions with human oversight.
• Figure 5: Demographic overview of survey respondents in both waves.