A new Reuters/Ipsos poll finds that 73% of Americans believe AI companies are not doing enough to prevent serious harm. The survey of 1,277 US adults also found that 39% think AI has a negative impact on society, while only 11% see it as positive. Most respondents want federal safety standards, and 55% support slowing AI development. The results came shortly after a cybersecurity incident involving OpenAI this summer.
Key Takeaways
- 73% of Americans believe AI companies lack strong safeguards against serious harm
- 39% of respondents view AI as harmful to society, up from 36% the previous month
- 55% of those polled support slowing AI development
- OpenAI experienced a cybersecurity incident this summer
- Nvidia CEO Jensen Huang rejects AI doomsday warnings, calling them unscientific
- Huang told CBS there is a 0% chance that 2030 will be the end of the world
- Trump proposed renaming AI to super intelligence in a UN address
- KT and NICE expand Korean AICE AI training to Vietnam, Sri Lanka, and Georgia
- Fort Worth police use a UTA AI tool to practice de-escalation techniques
- 44% of GCC companies report AI workforce talent shortages
Most Americans worry AI companies are not keeping society safe
A Reuters/Ipsos poll found that 73% of Americans believe AI companies are not doing enough to prevent serious harm. The four-day survey of 1,277 US adults also found 39% think AI has a negative impact on society, up from 36% last month. Most respondents want federal officials to set safety standards, and 55% support slowing AI development. The poll had a margin of error of 3 percentage points.
Survey shows Americans doubt AI safety efforts
A new Reuters/Ipsos survey found 73% of Americans do not believe AI companies have strong safeguards against serious harm. Thirty-nine percent said AI hurts society, while just 11% see it as positive. The poll also found 55% of people favor slowing AI development. The findings came after warnings from AI researchers and a cybersecurity incident involving OpenAI this summer.
Nvidia CEO Huang rejects AI doomsday warnings
Nvidia CEO Jensen Huang said claims that AI will end the world are irresponsible and not based on science. He told CBS there is a 0% chance that 2030 will be the end of the world. Huang said AI companies should focus on engineering safety and follow existing laws instead of asking for new rules. He also argued that fear-based warnings may hide other motives.
Santa Cruz residents share mixed views on AI
Local residents in Santa Cruz expressed a range of opinions about AI in a September 2026 street talk feature. Some, like barista Kate Motroni, distrust tech leaders and worry about data centers. Others, like business owner Dan Deppen, use AI daily and see it as useful. Several residents raised concerns about regulation, job loss, and AI making up false information.
AI changes consulting value toward expertise and execution
A new report from the Deerborne Group says AI is changing management consulting by making analysis faster and cheaper. A pre-ASCO 2026 survey found 62% of diagnostics and life sciences executives saw less analytical work, but 54% saw no clear improvement in decision quality. The report says consulting will add value through domain expertise, judgment, and execution rather than basic research.
AI panel experts forecast job shifts and product challenges in Kansas City
AI experts gathered at Back2KC in Kansas City to discuss the future of artificial intelligence. Julia Kanter said AI makes building prototypes faster but creates unrealistic expectations about how quickly they become real products. Ryan Shreeve of TENEX.AI said implementing AI requires many people and careful planning. Panelists said AI will change knowledge worker jobs through more reviewing rather than full replacement. They also warned that AI-powered cyberattacks will require equally advanced cybersecurity tools to fight back.
Middle East struggles to find skilled AI workers despite big investments
The Middle East has invested heavily in AI infrastructure like datacenters and GPU systems. However, 44% of GCC companies report workforce talent shortages as they move from AI experiments to full production. The biggest gaps include experienced AI engineers, data engineers, MLOps specialists, and people who can connect AI systems to real business needs. Experts say the shortage is worst for workers who can deploy models at scale and manage them in live environments. Some healthcare providers are training existing staff to help close the gap.
Trump proposes calling AI super intelligence at UN address
President Donald Trump addressed the United Nations General Assembly in New York on Tuesday. He proposed renaming artificial intelligence to super intelligence in all future US government documents. He expressed hope that other countries would adopt the new term. Later that day, Trump was scheduled to sign a security agreement covering Greenland with Danish Prime Minister Mette Frederiksen and Greenland Prime Minister Jens-Frederik Nielsen. The agreement grants the United States permanent control over security needs in Greenland.
KT and NIA expand Korean AICE AI training to developing countries
KT and the National Information Agency are expanding Korea's AICE AI training program to developing countries. The pilot training will launch this year in Vietnam, Sri Lanka, and Georgia. The program covers AI fundamentals, machine learning, and deep learning. It aims to build local AI capabilities that can help these countries in healthcare, finance, and education. The training will be conducted in partnership with local institutions and include collaboration with international experts.
Fort Worth police use UTA AI tool to practice de-escalation techniques
Fort Worth police officers are training with a new AI tool developed by the University of Texas at Arlington. The tool lets officers practice interactions with the public during traffic stops and other situations. Dr. Shuchi Deb designed the system with different difficulty levels, from cooperative to very challenging scenarios. Officer Jorge Lopez helped develop the training and said AI removes the need for physical human role players. The tool allows officers to practice more often and safely.
Sources
- Three out of four Americans say AI firms not doing enough to prevent disaster, Reuters/Ipsos poll finds
- 73 percent say AI firms not doing enough to prevent disaster: Survey
- ‘They must be doing it for ulterior reasons’: Jensen Huang says AI leaders are ‘irresponsible’ for scaring the public with ‘doomsday narratives’
- What Santa Cruz Locals Think About AI
- As AI Reshapes Consulting, Expertise and Execution Become the Differentiators
- AI makes it easy to start, but that’s not the hard part; tech builders forecast job shifts, product hiccups
- The Middle East’s AI race has a people problem
- Trump Tells UN Assembly AI Should Be Called 'Super Intelligence,' Signs Greenland Security Pact
- KT, NIA expand Korea’s AICE AI training to developing countries
- Fort Worth police learn de-escalation techniques with UTA AI tool
Comments
Please log in to post a comment.