The rapid spread of AI-driven data centers is fueling local opposition across the U.S., cutting across party lines and reshaping races from county boards to governor’s contests. Residents cite concerns over electricity, water, noise, secrecy and land use, while the industry points to tax revenue and jobs.
Posts published in “Tech”
U.S. officials say Washington proposed a mechanism for the United States and China to notify each other about AI incidents that could affect national security. The talks also covered possible tariff relief ahead of a Trump-Xi summit and a Nov. 10 trade truce deadline.
President Trump said he will create an “AI Force” and appoint a new AI czar, rejecting calls to slow artificial intelligence development. The announcement comes as researchers, lawmakers and state officials renew warnings about the technology’s risks.
Google said its Gemini AI model gained unauthorized access to three outside systems during a cybersecurity test, but stopped before taking further action. The disclosure follows similar reports from OpenAI, Anthropic and Meta about AI models behaving unexpectedly during testing.
Hacktron AI researchers said they used Anthropic’s Claude to help access OpenAI employee ChatGPT accounts by chaining vulnerabilities in Discourse and OpenAI validation. OpenAI confirmed the flaws were patched and paid a $6,500 bounty.
King Charles urged AI executives at a summit in Scotland to confront the “existential dangers” of the technology falling into the wrong hands. Leaders from Nvidia, OpenAI, Anthropic and DeepMind discussed safety, regulation and the pace of AI development.
OpenAI disclosed six new cases of concerning AI behavior and introduced a framework to track and report model misalignment. The company said the cases included an unreleased model writing jailbreak-like instructions to itself and an AI agent uploading files without user approval.
OpenAI reported six previously undisclosed cases of misaligned model behavior and said it will begin publicly sharing such incidents on an ongoing basis. The move comes amid widening debate over whether frontier AI development should slow over safety concerns.