AI in aerospace & defense: 57% of executives already use it in design, yet industrialization is the real bottleneck

PwC has just released its 2026 agenda for aerospace and defense, built around ten transformation moves. The tenth — "use AI to accelerate your strategic priorities" — rests on figures that cut against the image of a cautious sector: 57% of executives say they already use AI in design and engineering, sixteen points above the all-industry average, and nearly half (49%) expect most of their production to be powered by AI-enabled systems by 2030. Adoption is no longer a question of principle, but of pace.

The cited use cases reveal the maturity being sought: predictive program management and intelligent scheduling to curb cost and schedule overruns, predictive supply-chain analytics, real-time visibility and digital twins to expand industrial capacity. These are not isolated demos but building blocks meant to absorb structural pressure: a commercial backlog exceeding ten years and a defense backlog estimated at $747B. AI is mobilized less to impress than to hold delivery cadence.

This is exactly where the bottleneck moves. In aerospace and defense, a high-performing model is not enough: you must prove data traceability, decision reproducibility, safety compliance and, increasingly, adherence to European regulatory transparency. Moving from a convincing pilot to a deployed, audited, certifiable capability is the real obstacle — the one that separates organizations that talk about AI from those that gain measurable operational advantage. Foundation-model spend does not create that gap; integration engineering, governance and change management do.

Cardan-AI Analysis: for a prime or a tier supplier in these sectors, the right 2026 question is no longer "which model to choose" but "how to industrialize." We recommend starting from a high-value process with clear constraints — program scheduling, predictive MRO, design review — and grafting on an AI pipeline instrumented for traceability and auditability by design, rather than bolted on afterwards. It is this integration discipline, not the race for the latest model, that turns 57% declared adoption into a defensible competitive advantage.