Key Takeaways
- A Politico‑commissioned survey of six Western nations found that roughly half of respondents favor slowing AI development to curb risks.
- Between 55% and 65% of participants view the risk of AI‑caused human extinction as at least moderate, while only 22%‑33% consider it negligible.
- Political affiliation shows little influence on AI‑risk perceptions; 44% of 2024 Trump voters still support slowing AI, compared with 40% who prioritize staying ahead technologically.
- UN Secretary‑General António Guterres has urged the U.S. and China to open a dialogue on AI governance, while former President Trump announced plans for dedicated AI units and an “AI czar.”
- Despite widespread concern, a sizable minority (30.6%‑39.8%) advocates for pushing forward with more advanced AI models to reap potential benefits.
Survey Overview and Methodology
The poll, commissioned by Politico and fielded across the United Kingdom, Germany, Spain, Canada, the United States, and France, sought to gauge public sentiment toward the trajectory of artificial intelligence. Researchers asked respondents whether they believed AI development should be slowed to mitigate risks, accelerated to capture benefits, or left unchanged. The survey captured a broad demographic cross‑section, allowing analysts to compare attitudes by nation and political leaning. As the report notes, “About half of the respondents in a survey commissioned by Politico supported slowing the development of artificial intelligence to mitigate potential risks.” This foundational figure set the tone for the ensuing analysis of risk perception and policy preference.
Perceived Extinction Risk
When asked to evaluate the likelihood that AI could lead to humanity’s destruction, a clear majority expressed concern. On average, 12% of participants across the six countries judged the risk “almost certain,” while another 25.8% rated it as “moderate.” Combining these categories yields a range of 55%‑65% of respondents who see the threat as at least moderate, depending on the nation. Conversely, only 22%‑33% assessed the danger as “very low” or dismissed it entirely. The survey quotation underscores this split: “The proportion of people who consider the risk of humanity’s destruction due to AI to be significant—ranging from moderate to nearly inevitable—ranged from 55% to 65%, depending on the country.” This data highlights a pervasive unease that transcends borders.
Support for Slowing AI Development
Reflecting the risk concerns, support for decelerating AI progress was substantial but not unanimous. Across the six nations, the share of respondents who believed AI is already sufficiently advanced and advocated for slowing further development fell between 45.1% and 50.8%. This narrow band indicates that roughly one‑in‑two citizens favor a precautionary approach. The report frames this sentiment succinctly: “Overall, the proportion of those who believe artificial intelligence is already sufficiently advanced and support slowing its further development ranged from 45.1% to 50.8% across the six countries.” Such a figure suggests a strong, though not overwhelming, appetite for regulatory restraint.
Advocacy for Accelerated AI Innovation
Opposing the cautionary stance, a notable minority pushed for advancing AI capabilities. Between 30.6% and 39.8% of respondents endorsed creating more advanced models, arguing that the potential benefits outweigh the risks. This group often emphasized economic competitiveness, scientific breakthroughs, and societal improvements as motivations for rapid development. The survey’s wording captures this perspective: “At the same time, between 30.6% and 39.8% of respondents supported the creation of more advanced models that could bring significant benefits, while the rest were undecided.” The presence of a sizable pro‑innovation bloc underscores the policy tension between safety and progress.
Political Ideology and AI Attitudes
One of the survey’s most striking findings was the limited impact of political affiliation on AI risk perception. Even among voters who backed Donald Trump in the 2024 U.S. election, opinions were split: 44% favored slowing AI development, while 40% argued that the United States should not fall behind other nations, even if it entails accepting additional risk. The report states, “Specifically, among those who voted for Donald Trump in 2024, 44% believe that the development of artificial intelligence should be slowed down, while 40% do not support this idea and believe the U.S. should not fall behind other countries, even if it means taking on additional risks.” This near‑even division suggests that concerns about AI safety cut across partisan lines, challenging the notion that technology policy is strictly a liberal‑conservative issue.
International Calls for AI Governance
Beyond domestic opinion, the survey coincided with heightened international discourse on AI governance. UN Secretary‑General António Guterres publicly urged the United States and China to engage in dialogue aimed at establishing common norms and safeguards for artificial intelligence. His appeal reflects growing multilateral anxiety that unchecked AI advancement could destabilize global security. Simultaneously, former President Trump announced plans to create dedicated AI units within the federal government and to appoint an “AI czar” to oversee national strategy. These moves signal a recognition, at the highest levels, that AI demands coordinated policy attention—both domestically and on the world stage.
Balancing Risk and Opportunity
The survey data collectively portray a public that is simultaneously wary of AI’s existential hazards and intrigued by its transformative promise. While a majority see at least a moderate risk of catastrophe, a significant minority remains eager to push the frontier of machine intelligence in hopes of unlocking breakthroughs in medicine, climate modeling, and economic productivity. This dichotomy poses a classic policy dilemma: how to impose sufficient safeguards without stifling innovation. As the Politico report encapsulates, “Overall, the proportion of those who believe artificial intelligence is already sufficiently advanced and support slowing its further development ranged from 45.1% to 50.8% across the six countries. At the same time, between 30.6% and 39.8% of respondents supported the creation of more advanced models that could bring significant benefits…” The challenge for policymakers lies in crafting frameworks that address the legitimate fears of the former group while nurturing the aspirations of the latter.
Conclusion
The Politico‑sponsored survey offers a nuanced snapshot of Western attitudes toward artificial intelligence. Roughly half of respondents favor slowing AI development to mitigate risks, while a comparable share perceives the threat of AI‑driven human extinction as moderate to near‑certain. Political leanings appear to have minimal sway over these views, with even Trump supporters split between caution and competitiveness. International voices—ranging from the UN Secretary‑General to national leaders—are increasingly calling for dialogue and structured governance to navigate AI’s dual nature. As nations grapple with how to balance safety and progress, the survey’s findings underscore the need for inclusive, evidence‑based policies that reflect both the apprehensions and ambitions of their citizens.
https://ua.news/en/world/bilshist-zhiteliv-zakhodu-poboiuietsia-shcho-shtuchnii-intelekt-mozhe-znishchiti-liudstvo