The biggest unresolved question in AI right now, and more takeaways from Madrona’s IA40 Summit
Who owns the customer when an AI agent stands between them? Microsoft, Amazon, Anthropic and Stripe can't agree — and it's blocking real deployments.
Why it matters
Enterprise AI agents are moving from pilots to production, but a fundamental governance gap remains unsolved: when an agent mediates a customer interaction, who retains the relationship, the data rights, and the compliance burden? The IA40 Summit exposed deep disagreement among major vendors, revealing that most enterprises are far behind both the technical and contractual maturity Anthropic has achieved internally.
The key facts
5 to knowPanelists from Microsoft, Amazon, Anthropic, and Stripe discussed agent-mediated customer interactions without resolving data ownership or customer relationship accountability
Most companies are significantly behind Anthropic's own AI gains — suggests a gap between lab capability and enterprise deployment velocity
Divergence on vendor lock-in: split opinion over relying on a single AI provider
Data governance and identity boundaries remain unresolved in agent-to-business-to-customer workflows
No pricing, SLA, or contractual models disclosed for agent-mediated customer interactions
Go to the source
GeekWiregeekwire.com
Publisher excerpt: Executives from Microsoft, Amazon, Anthropic and Stripe described AI agents stepping between companies and their customers, without settling who keeps the customer relationship or the data that comes from it. Also: why most companies are far behind Anthropic's own gains from AI, and a split over…