OpenAI safety researcher David Robinson (opens in a new tab) resigned this week, warning that the company's culture is broken (opens in a new tab) as speed trumps safety. His departure follows a series of high-profile exits and comes as Anthropic meets religious scholars (opens in a new tab) to shape AI morality, while Meta's Muse (opens in a new tab) faces scrutiny over its personal profiles.

Meanwhile, the industry's shift from chatbots to autonomous agents accelerates. Developers argue agents need documentation, not memory (opens in a new tab), echoing wider debates about reliability and safety. The controversy intensifies as Yann LeCun dismisses extinction fears (opens in a new tab) just as OpenAI's internal strife reignites questions about accountability.