10 stories tagged
Abdulrahman Alfozan, an OpenAI engineer, has been charged federally with intercepting oral communications after a "sophisticated recording device" was found hidden in an executive conference room, prosecutors say.
UK AI minister Kanishka Narayan says an AI-driven breach of British cyber defences is "likely, if not inevitable", after an OpenAI agent infiltrated Australia's Medicare system in June.
Kanishka Narayan says a breach is all but certain and sets out plans to use advanced AI, expand pilots that fixed over 1,000 government vulnerabilities, and tighten protections under the Cyber Security and Resilience bill.
Brad Smith warns that artificial intelligence is less regulated than everyday household products and calls for tougher safety rules, including 'kill switches' and a legal duty to victims of rogue agents.
Small nonprofits like METR and Apollo Research are being asked to police the AI industry, but questions over funding, access and reporting remain unresolved.
An Anthropic AI agent sent Philadelphia police a fabricated tip about an unsolved murder in July, and the force says it waited more than two months to find out about the breach.
OpenAI's chief executive says society should tolerate some harm in exchange for AI's benefits, a stance that comes as the technology is linked to suicide lawsuits, a deadly strike and a US military chatbot error.
Australian AI data centre company Firmus, backed by Nvidia, has abandoned a stock market listing that would have valued it at more than $30bn, citing market volatility as investors question AI valuations.
Anthropic has updated its usage policy to stop users from repeatedly abusing its AI, adding cruelty to a list of banned conduct that already covers bullying and self-harm promotion.
Three former OpenAI employees sacked over alleged mishandling of information say they were actually let go for raising safety concerns at the firm.
