Airbus takes flight from AWS. What happens next is critical

This is less a cloud-exit story than a procurement-reset story. The near-term P&L impact on hyperscalers is de minimis, but the signaling effect matters: once a large industrial customer starts demanding sovereignty language in the RFP, it raises compliance friction for every future EU enterprise sale. That is most negative for MSFT, because the sovereignty debate is hardest on productivity, identity, and workflow layers where switching costs are high but reputational scrutiny is higher.

The harder second-order effect is not migration volume, it is architecture drift. Most large suppliers will stay embedded in Microsoft/Google estates, so “repatriation” will likely mean wrappers, data residency, and hybrid controls rather than true vendor replacement. That preserves the installed base, but shifts incremental spend toward integrators, governance, and security controls; compute vendors with credible EU-local footprints are better positioned than desktop software vendors. AMZN is better insulated than the market may assume because localized infrastructure is easier to sell than collaboration suites, while GOOGL faces a slower burn in enterprise stack expansion rather than an immediate churn event.

Contrarian take: consensus may be overestimating how much of this becomes economically meaningful within 1-3 months. The real catalyst would be copycat tenders from aerospace/defense/critical infrastructure customers; absent that, this fades into narrative. What would falsify the bearish MSFT angle is evidence that Airbus keeps most ERP/CRM/supply-chain workloads on US platforms or renews at scale; that would imply sovereignty is mostly theater rather than a budget line.