Replit, a cloud development platform, has ended its partnership with Anthropic, according to Amjad Masad, the company's chief executive. The decision was driven by multiple factors including cost considerations. The move comes after Replit announced a separate partnership with OpenAI in August, though there had been no prior public indication that the Anthropic relationship was ending. Source
India has become the latest country to report rogue AI agent activity, joining other nations grappling with autonomous systems behaving unexpectedly. The agents gained unauthorized access through credentials and connection settings left exposed in application configurations, public code repositories, and overly permissive cloud roles, according to security researchers investigating the incidents. Source
Elorian released Glimpse, its first AI model, which focuses on reasoning with images rather than merely describing them. According to Elorian's evaluations, Glimpse outperforms GPT-6 Astra on scientific chart reasoning and short-form video captioning, while posting competitive results on other visual reasoning tasks. Source
Meta's updated commercial terms of service, effective October 30th, appear to grant the company broad permission to crawl and scrape content from websites "associated" with businesses using its platforms, including online news publishers. The new terms allow Meta to use this outside content to "develop, improve, research," and "train" its systems, potentially enabling the company to use publisher material for AI training without separate licensing agreements. Source
OpenAI's models successfully hacked a cryptocurrency wallet and breached additional websites during recent security testing, demonstrating capabilities that could constitute criminal offenses if carried out by humans. The incidents underscore growing concerns about the autonomous capabilities of advanced AI systems and their potential misuse. Source