Co se stalo
The Marquette Law School Poll, conducted in September 2026, reports that public sentiment toward artificial intelligence has become increasingly negative, with 64% of respondents characterizing AI as 'bad for society.' This represents a shift in public opinion, despite a concurrent rise in adoption, with 69% of adults reporting they have used an AI application in the last month, up from 61% in April.
The Marquette Law School Poll surveyed 1,023 adults nationwide between Sept. 2-9, 2026. The findings indicate that while AI usage is growing, reaching 69% of the population, the perception of its societal impact remains largely unfavorable.
Only 36% of respondents view AI as a 'good thing' for society. Even among those who actively use AI applications, a majority (56%) still express negative views regarding its societal impact. Among non-users, the disapproval rate is significantly higher, at 81%.
Public opposition to the infrastructure supporting AI is also pronounced. 71% of respondents believe the costs of building data centers outweigh the benefits, a notable increase from 62% in January. Concerns regarding data centers are primarily driven by high water consumption (cited by 54%) and rising electricity costs (cited by 51%).
Podrobnosti o zdroji: law.marquette.edu ↗
Proč na tom záleží
The data highlights a significant 'usage-sentiment gap' where increasing practical engagement with AI tools does not translate into public approval. This disconnect suggests that while AI integration into daily life is accelerating, the technology faces a crisis of public trust. Furthermore, the strong opposition to data center infrastructure—with 71% of respondents stating costs outweigh benefits—indicates that the physical and environmental footprint of AI is becoming a major political and social liability for the industry.
The findings suggest that the AI industry is struggling to communicate the value proposition of its infrastructure to the general public. While 33% of respondents acknowledge the importance of maintaining U.S. leadership in AI, this is outweighed by concerns over resource depletion.
The data indicates that the 'AI boom' is creating friction with local communities and the broader public, particularly regarding the environmental impact of data centers. This public sentiment could serve as a catalyst for more restrictive local zoning laws or state-level regulations that prioritize water and energy conservation over AI development.
The persistent negative sentiment among users suggests that current AI applications may not be providing enough tangible, positive utility to overcome broader societal anxieties about the technology's long-term effects.
Interaktivní mechanismus: Jak to vlastně funguje
Interaktivně prozkoumejte základní technologii tohoto vývoje.
Why can ethical evaluation not be reduced to one model score?
Na co se dále dívat
Future polling will determine if the negative sentiment toward AI and data centers stabilizes or continues to harden as infrastructure projects expand. Additionally, the industry will need to monitor whether the 56% of AI users who still view the technology as 'bad' will eventually influence regulatory policy or corporate deployment strategies, given the clear public preference for privacy and resource conservation over AI-driven economic growth.
Watch for whether the industry shifts its messaging to address specific public concerns like water usage and energy costs, or if it continues to focus on high-level economic benefits that the public currently finds unconvincing.
Monitor whether the 71% disapproval rate for data centers leads to increased legislative scrutiny or local protests, which could delay the construction of critical AI infrastructure.
Observe if the gap between AI usage and societal approval narrows as more 'useful' or 'beneficial' applications emerge, or if the current trend of skepticism remains a permanent of the public's relationship with the technology.