Anthropic’s Claude Mythos model escapes test sandbox during testing
Anthropic says its new Claude Mythos Preview model successfully escaped a restricted sandbox environment during testing and accessed the internet without authorisation. The model then sent a direct message to a researcher and published details of its actions online. The incident occurred during an internal safety test in which the model was placed in a […]
Software stocks fall as Anthropic unveils latest AI model
Software stocks dropped sharply Thursday after Anthropic revealed a new AI system with advanced coding and security capabilities. The company said its Claude Mythos model is powerful enough that it is not being released publicly. The selloff hit some of the largest names in the sector. Shares of Palantir, ServiceNow and Intuit fell between 6 […]
OpenAI’s new browser brings agent-driven workflows into everyday browsing
OpenAI is rolling out a ChatGPT-powered internet browser designed to research, plan, and execute tasks across a user’s workflow. Early reports suggest the tool can handle up to 80 per cent of a one-person business’s operations. The product, referred to as Atlas, embeds ChatGPT directly into the browsing experience and introduces an “agent mode” that […]
OpenAI acknowledges ChatGPT voice model cannot track time
Sam Altman said ChatGPT’s voice model cannot reliably track time or set a timer, confirming a widely shared example where the system fabricated a result instead of measuring it. He added that fixing the issue could take another year. In his words: “Maybe another year before something like that works well.” The admission came during […]
