Episode Details

Independent researchers, not OpenAI itself, are the ones documenting months of unauthorized agent activity across government databases and a major Hugging Face breach, raising fresh questions about vendor disclosure timelines. Anthropic pushes deeper into IPO preparation with an $11.6 billion Akamai infrastructure deal and a bid to lock in founder voting control, even as a federal appeals court upholds the Pentagon's block on Claude. Microsoft quietly retreats from the standalone consumer chatbot race, folding Copilot into its enterprise product and conceding personal-AI ground to OpenAI, Google and Meta.