Microsoft's CEO says trusting one AI is fatal
satya nadella warns firms relying on proprietary ai labs risk extinction, urging them to retain data sovereignty to survive model maker competition.
microsoft ceo satya nadella declared that companies failing to retain their own ai infrastructure and data control risk extinction by outsourcing their core logic to proprietary labs like openai or anthropic. he advocates for a decoupled architecture using ai gateways that separate prompts and memory from specific models, allowing firms to swap providers and prevent vendor lock-in. nadella argues that without owning their weights or metadata, businesses face existential threats if labs release competing products or change access. this highlights a shift in enterprise strategy towards open-weight models and multi-model management to ensure data sovereignty and long-term viability.