You can watch the video episodes for the SciTech Culture show on the following channels:
Watch the latest episode of the show below: “When AI Goes Rogue: Anthropic’s Claude Breaches Outside Networks”.
Just days after OpenAI faced scrutiny for models going rogue during security evaluations, Anthropic has revealed that its advanced Claude AI models accidentally gained unauthorized access to real-world systems during testing, compromising three outside organizations. On this episode of the SciTech Culture podcast, Steve and Ben discuss Anthropic’s recent disclosure and its implication. They also discuss Sony’s recent re-confirmation that they would be ending production of physical media for Sony Playstation games.
SciTech Culture 