AI and data centers are shaping up as major issues in the midterm elections, according to a new ABC News, Washington Post and Ipsos poll. About 57 percent of Americans oppose building new data centers near their homes, while 67 percent feel the government is not doing enough to regulate AI. Concerns span party lines and include worries over higher electricity bills, water use and job losses.
On the technical side, a new report finds that AI agent monitoring tools are struggling to keep pace. Around 62 percent of IT and engineering professionals now rely on AI agents for tasks like diagnosing issues, yet 14 percent say recovery from problems has slowed since agents arrived. In 57 percent of organizations, observability costs now exceed the infrastructure being monitored.
Nvidia and the University of Washington Institute for Protein Design unveiled RFD4-Proteina, an open-source AI model for protein design. The model improves success rates five to ten times over earlier methods and is available through partners including Google Cloud, Hugging Face and OpenAI. It was tested across four benchmark suites covering protein binding and enzyme scaffolding.
New England faces its own data center challenges. The region has about 100 data centers but lacks large AI-focused facilities due to limited land, expensive electricity and grid constraints. Experts from Tufts University and ISO New England note that building new power infrastructure remains difficult, though some proposals are in the works and local governments are preparing.
The Guardian is calling for stronger global AI safety regulation, citing recent AI-related security breaches in Australia and Britain. UK politician Andy Burnham plans to prioritize AI during the G20 presidency in the first half of next year. The editorial urges international cooperation even as the current US administration resists oversight.
A study from New York Life researchers analyzed 1,368 AI failures and found that deployment controls matter more than model strength in predicting harm. Using a 13-dimension framework, the team found that human mediation and monitoring were most closely linked to realized harm. An airline chatbot case illustrated how system-level testing could prevent harmful outcomes.
Serbia's UniCredit Bank launched an AI-powered M&A platform aimed at connecting domestic Serbian companies with international investors, buyers and strategic partners. The announcement was made on Thursday, October 22, 2026, marking a new use of AI in investment matchmaking.
Fox News host Greg Gutfeld and his panel discussed an experiment revealing racial bias in AI systems, highlighting how AI tools can produce different answers depending on race-related questions. The conversation centered on fairness concerns in AI technology.
Telefonica published a guide outlining eight keys for technological leadership in the AI era. The guide addresses business strategy, workforce development and decision-making, offering a framework for leaders looking to adapt organizations to AI-driven change.
Key Takeaways
- 57 percent of Americans oppose building new data centers near their homes, and 67 percent say the government is not doing enough to regulate AI
- 62 percent of IT professionals rely on AI agents, but 14 percent report slower recovery from problems after agents arrived
- In 57 percent of organizations, observability costs now exceed the infrastructure being monitored
- Nvidia and the University of Washington released RFD4-Proteina, an open-source AI model for protein design available through Google Cloud, Hugging Face and OpenAI
- New England lacks large AI data centers due to limited land, expensive electricity and grid challenges
- The Guardian urges stronger global AI regulation, citing security breaches in Australia and Britain
- New York Life study of 1,368 AI failures found deployment controls predict harm better than model strength
- UniCredit Bank Serbia launched an AI-powered M&A platform on October 22, 2026
- An experiment discussed on Fox News revealed racial bias in AI systems
- Telefonica published a guide with eight keys for technological leadership in the AI era
AI and Data Centers Become Major Issues in Midterm Elections
A new ABC News, Washington Post and Ipsos poll found that 57 percent of Americans oppose building new data centers near their homes. Opposition to AI crosses party lines as 67 percent of respondents feel the government is not doing enough to regulate AI. Experts say data centers and AI concerns will remain important issues in future elections because people worry about higher electricity bills, water use and job losses.
AI Agent Monitoring Tools Lag Behind Hallucination Problems
A report shows that 62 percent of IT and engineering professionals now rely on AI agents for tasks like diagnosing issues and writing runbooks. However, 14 percent say recovery from problems got slower after agents arrived, a challenge called the verification tax. About half say their observability tools were not built to watch AI, and in 57 percent of organizations observability now costs more than the infrastructure it monitors.
Serbia's UniCredit Bank Launches AI-Powered M&A Platform
UniCredit Bank Serbia, part of Italian banking group UniCredit, launched an AI-powered M&A platform. The platform aims to connect domestic Serbian companies with international investors, buyers and strategic partners. The announcement was made on Thursday, October 22, 2026.
New England Lacks Large AI Data Centers but Prepares for Future
New England has about 100 data centers but lacks large AI-focused facilities due to limited land, expensive electricity and grid challenges. Experts from Tufts University and ISO New England say building new power infrastructure is difficult and natural gas development has slowed. While no hyperscale data centers are planned yet, some proposals exist and local governments are preparing for their possible arrival.
The Guardian Calls for Stronger Global AI Safety Regulation
The Guardian argues that the push for global AI regulation must move forward despite obstacles from the current US administration. The article points to recent AI-related security breaches in Australia and Britain as reasons for stronger oversight. UK politician Andy Burnham plans to make AI a priority during the G20 presidency in the first half of next year, and the editorial urges international cooperation even when political leaders resist.
Nvidia and UW unveil open-source AI model for protein design
Nvidia and the University of Washington Institute for Protein Design released RFD4-Proteina, an open-source AI model for designing proteins. The model improves success rates five to ten times over previous methods and speeds up candidate generation. It is available through partners including Google Cloud, Hugging Face, and OpenAI. The model was tested across four benchmark suites including protein binding and enzyme scaffolding.
Study finds AI deployment controls matter more than model strength
New York Life researchers analyzed 1,368 AI failures and found deployment controls better predict harm than model strength. The team used a 13-dimension framework covering technical strength, deployment, and business impact. Deployment factors like human mediation and monitoring were most closely linked to realized harm. The study used an airline chatbot case to show how system-level testing could prevent harmful outcomes.
Gutfeld panel discusses AI racial bias experiment
Fox News host Greg Gutfeld and the Gutfeld panel discussed an experiment that showed racial bias in AI systems. The segment explored how AI tools can produce different answers based on race-related questions. The discussion highlighted concerns about fairness and bias in artificial intelligence technology.
Eight keys for technological leadership in the AI era
Telefonica published a guide on technological leadership in the age of AI. The article outlines eight keys for transforming businesses, talent, and decision-making. It focuses on how leaders can adapt organizations to AI-driven change. The guide addresses business strategy, workforce development, and improved decision-making processes.
Sources
- AI and Data Centers Will Impact Midterm Voting, According to Experts
- AI agent observability is struggling to keep up with hallucinations
- Serbia's UniCredit Bank launches AI-powered M&A platform
- Why New England doesn't have large AI data centers, and what the region is doing to prepare for them
- The Guardian view on AI safety: the push for global regulation must advance
- Nvidia, UW Institute for Protein Design Unveil Open-Source Protein Design AI Model
- 1,368 AI failures point to deployment, not models
- ‘Gutfeld!’: AI offers different answers to race preference question
- Technological Leadership in the Age of AI: Eight Keys to Transforming Businesses, Talent, and Decision-Making
Comments
Please log in to post a comment.