Executive Summary
The majority of Americans are worried about the consequences of AI for their safety and American politics. However, concerns about AI varies considerably among partisan groups, races, age groups, and household incomes.
Key findings include:
- Worries about AI are widespread:
- 49.8% of Americans expect AI to have negative consequences for the safety of elections.
- 65.1% of Americans are worried that AI will harm personal privacy.
- 40.0% of Americans think that AI will make national security worse.
- 38.3% of Americans are concerned about the use of AI in law enforcement.
- 50.1% of Americans anticipate elections becoming less civil due to the use of AI.
- 49.6% of Americans are worried that AI will threaten their jobs.
- Many Americans are unsure about the implications of AI: Between 17% and 36% of Americans do not have a clear opinion about the consequences of AI for safety.
- Republicans are slightly more concerned than Democrats: Across all issues, Republicans tend to be more worried about the consequences of AI than Democrats, but not by much.
- Younger, more affluent, and non-white Americans are less worried about AI: Although skepticism about AI is common across all sociodemographic groups, younger generations are less worried about the implications of AI than older ones. Furthermore, White Americans tend to have greater concerns about AI than non-White Americans. Likewise, more affluent Americans tend to be more optimistic about the rise of AI.
Main Results
This the third in a series of monthly reports on the state of partisan animosity conducted by the Polarization Research Lab (PRL) at Dartmouth College, University of Pennsylvania, and Stanford University. The focus of this March 2024 report is artificial intelligence (AI).
Over the last 16 months we have tracked American partisan attitudes with 74,354 survey interviews on the YouGov platform.
This report looks at trends across this full period and specifically 2,919 interviews conducted in March 2024.
Americans are worried about the implications of AI
We asked our respondents if they think AI makes national security, personal privacy, and American elections better or worse.
Americans are worried about the consequences of AI in law enforcement and elections
We also asked respondents more specifically about the consequences of AI for law enforcement and electoral campaigns. Concerning the consequences for law enforcement, we informed respondents that AI is already used to help set bail conditions, identify those who are likely to commit crimes again, and to identify tax cheats. We furthermore inform respondents that the use of AI in law enforcement is expected to grow dramatically in the future.
A majority of Americans fear AI might threaten their jobs
Lastly, we asked our respondents whether they are worried about their jobs due to the rise of AI.
Republicans are slighly more concerned about AI than Democrats
Evaluations of the benefits of AI may differ along party lines. Indeed, Republicans are more concerned about the implications of AI for personal privacy, national security, and elections. While Democrats are also worried about the effects of AI on personal privacy and elections, they are divided over whether AI improves national security. Independents are more pessimistic about the consequences of AI for all three areas, especially personal privacy.
Across all partisan camps, a majority of Americans expect the use of AI in electoral campaigns to make elections less civil. While a majority of Democrats and Republicans thinks that AI will make law enforcement less fair, just under a third of Democrats and Republicans considers AI to make the criminal justice system more fair.
A majority within all partisan groups is worries that AI will threaten their jobs, with Republicans being slightly more concerned than Democrats.
Racial Differences and AI
In 2024, AI-generated images surfaced on social media depicting Republican candidate Donald Trump surrounded by a group of black voters. In response, observers raised concerns that AI may accelerate the spread of misinformation and target different communities with fabricated imagery and information (source).
Nevertheless, white Americans are the most concerned about the consequences of AI, while Black voters are the least concerned. Concerning AI threatening respondents’ jobs, White and Black Americans express similar levels of concern about job security, with Asian Americans being the most concerned.
Younger Americans are much more optimistic about the use of AI
Technological change often makes older generations feel threatened. Our survey data is consistent with this proposition: Across all items asked in our survey (except for job insecurity), younger respondents tend to evaluate the consequences of AI more positively than older ones.
Lower-income Americans slightly more worried about AI
The rise of AI may affect Americans to different degrees depending on their income, and might thus lead to diverging evaluations of AI. Americans with a lower income (household income less than $100,000) are slightly more concerned about the consequences for personal privacy, national security, and elections. Similarly, less affluent Americans tend to have greater concerns about the role of AI in law enforcement and electoral campaigns and job security.
Appendix
About The Polarization Research Lab and our Data
- The Polarization Research Lab works to understand and strengthen democracy by conducting rigorous science, producing public goods, and training the next generation of scholars. The Lab is led by PIs Sean J. Westwood at Dartmouth College, Yphtach Lelkes at the Annenberg School for Communication at University of Pennsylvania, and Shanto Iyengar at Stanford University.
- PRL studies the political attitudes of Americans and the behavior of elected officials. Our goals are to:
- Dispel rumors and show hard data on the democratic attitudes of citizens. We conduct the largest continuous tracking poll on YouGov, collecting 1,000 interviews of Americans a week.
- Identify the behavior of elected officials that contributes to toxic polarization.
- Produce publicly available reports and tools to help stakeholders advance responses to toxic polarization that are based in data and evidence. PRL works directly with journalists, democracy practitioners, and policymakers.
PRL Is Supported by
- The Charles Koch Foundation
- The Hewlett Foundation
- The Knight Foundation
- The Templeton World Charity Foundation
- New Pluralists
Survey Questions: AI
Do you think artificial intelligence will make each of the following better or worse? In a grid: National Security, Personal Privacy, American Elections
- Better
- Worse
- Don’t know
Artificial intelligence is currently used to help set bail conditions, identify those who are likely to commit crimes again, and to identify tax cheats. This is expected to grow over the coming years. Do you think artificial intelligence will make law enforcement more or less fair in America?
- A lot more fair
- More fair
- Neither more nor less fair
- Less fair
- A lot less fair
If artificial intelligence is used by politicians during elections, do you think elections will become more or less civil?
- A lot more civil
- More civil
- Neither more nor less civil
- Less civil
- A lot less civil
Do you agree or disagree: over the next 20 years, the use of artificial intelligence in the workplace will threaten your job.
- Strongly agree
- Agree
- Neither agree nor disagree
- Disagree
- Strongly disagree
Additional data on American attitudes is publicly available at PRL’s America’s Political Pulse.
Copyright 2024 Polarization Research Lab