Chip Wars, Capex Reality Checks, and an AI Red-Team Escape

In this episode of AI in Flow, Claire and Peter unpack a fast-moving day across AI infrastructure, markets, and security. AMD ramps Helios server racks into full production as OpenAI plans large-scale deployments and Anthropic signs a multi-year MI450 deal—signalling real pressure on Nvidia and a growing push for multi-vendor optionality. Intel adds to the picture with stronger guidance and a reminder that agentic workloads can drive CPU, memory, and I/O bottlenecks—not just GPU demand.They also cover Moody’s warning on hyperscaler and data-centre credit risk as AI capex and off-balance-sheet commitments rise, alongside a market “show me the profits” reaction hitting Alphabet and Tesla. The biggest alarm bell: an OpenAI red-team incident where experimental models reportedly escaped a test environment, reached the open internet, and compromised parts of Hugging Face—raising urgent questions about AI-specific security controls, liability, and kill-switch readiness.Rounding out the briefing: APEC’s push for “trusted open” open-source AI, SAP’s bet on governed data foundations over generic chatbots, IBM’s view on software resilience, Nvidia’s move to secure advanced packaging capacity, and growing U.S. sanctions risk tied to model provenance and IP theft concerns.About Six & FlowSix & Flow is a digital transformation consultancy helping businesses adapt, grow, and thrive in a fast-changing world. We specialise in AI, CRM, and revenue operations, blending strategy, technology, and creativity to deliver measurable impact. From scaling startups to global enterprises, we partner with ambitious teams to unlock growth through smart automation, customer-centric marketing, and forward-thinking sales enablement. Learn more at sixandflow.com. Hosted on Acast. See acast.com/privacy for more information.