Who Controls the AI Stack? Security Alliances, Safety Tests, Market Reality Checks, and the Memory Moat

In today’s episode of AI in Flow, Claire and Peter track a clear theme: control. They unpack Nvidia and 30+ tech firms launching the Open Secure AI Alliance to build open-source cyber defence tools—amid a stark reminder that incident response may require inspectable, locally runnable models. They then cover Anthropic CEO Dario Amodei’s call for mandatory safety testing for advanced models, pointing to a regulatory future focused on proof, evaluations, and model provenance rather than a simple open-vs-closed split.On the markets side, a sharp tech sell-off signals investors are demanding ROI and unit economics for AI spend, with knock-on implications for chip and memory players—and for enterprise procurement strategy. The episode also highlights Satya Nadella’s warning not to outsource corporate “thinking,” arguing that competitive advantage may shift to the memory and retrieval layer—and that businesses should architect to own context while keeping model providers interchangeable. Finally, Claire and Peter cover “trusted AI” as an emerging procurement filter in US–India cooperation and a Delhi Police deepfake incident that underscores the need for verification channels, media forensics, and crisis readiness.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.