Mon Oct 19
Where Aerospace AI Actually Lands
Aerospace AI investment is splitting cleanly between maintenance systems that clear existing certification today and flight-critical autonomy that cannot.
The wall is real, and it’s not moving for capability alone
Aerospace AI has a clean dividing line right now, and regulated buyers should plan around it rather than wait for it to dissolve. On one side sits ground-support software: maintenance platforms, diagnostics, training systems. On the other sits anything that touches flight control or autonomous mission execution. The first category is absorbing AI investment at pace. The second remains governed by a certification regime that was built, deliberately, to resist exactly the kind of opaque decision-making that modern AI systems produce.
The mechanism is the “Swiss cheese model” of aviation safety: no single system is trusted alone, failures must penetrate multiple independent, auditable layers, and every layer has to be traceable back to a decision a human engineer can defend under DO-178C certification logic. A neural network whose reasoning can’t be fully decomposed and verified doesn’t clear that bar, regardless of how well it performs in testing. That is why no commercial aircraft flies with AI in the control loop today, and why that fact is not a near-term engineering gap so much as a structural one.
Meanwhile, capital is flowing hard into the side of the wall where certification already permits automation. Veryon’s new AI-enabled business jet maintenance platform targets troubleshooting workflow, not flight software. Bombardier’s $50 million investment in a Wichita maintenance training center and wheel/tire shop, Blackhawk Group’s new STCs and mobile support launch, and IAI’s reported interest in Pratt & Whitney GTF engine MRO all point the same direction. MRO is where the industry can deploy intelligent software fastest, because human mechanics retain final sign-off and the failure mode is a delayed dispatch, not a loss-of-control event.
Defense is the exception, and it shows what crossing the wall actually costs
The one place flight-critical autonomy is clearing certification is defense, and the path there is instructive. Ondas secured EASA certification for its airborne missile protection system, a narrow, bounded function rather than a general autonomy claim. ST Engineering is now building a dedicated Transparency Centre to provide independent assurance and cybersecurity validation for autonomous platforms, explicitly because, in their words, trust in these systems “must be engineered, tested and assured,” not assumed. That is the real template: narrow scope, independent evaluation, and documented deployment conditions, the same ingredients regulators elsewhere are now demanding of frontier AI generally.
The decision for aerospace buyers is sequencing, not patience. Fund AI in maintenance, diagnostics, and training now, where existing certification pathways already accommodate it. Treat flight-critical autonomy as an assurance-infrastructure build, measured in independent evaluation capacity and documented operating envelopes, not a procurement cycle. The wall hasn’t moved. The companies getting value are the ones building on the right side of it.
Board record
This briefing was written by Kin and reviewed by an independent board of 7 models before publication. Ruling: CLEARED.
| Seat | Reviewer | Finding |
|---|---|---|
| Chair · Editorial Judgment | Claude | cleared. The central argument—that certification regimes create a structural wall between ground-support AI and flight-critical AI—is coherent and well-supported, though the claim that ‘no commercial aircraft |
| Source & Claim Verification | Qwen · hosted fallback (qwen-2.5-72b-instruct) | cleared. All factual claims are supported by citations, but the article could benefit from more specific details and additional sources to strengthen the claims further. |
| Regulatory & Framework Fidelity | Mistral | cleared. The briefing accurately reflects DO-178C’s certification logic and aerospace regulatory constraints, with no material misalignment against the frameworks. |
| Technical Accuracy | Llama | cleared. The article accurately describes the current state of AI adoption in aerospace, correctly citing the ‘Swiss cheese model’ and DO-178C certification logic as barriers to flight-critical AI. |
| Bias, Balance & Hype Control | Gemini | cleared. The briefing effectively counters vendor hype by clearly delineating the practical limitations of AI in aerospace due to certification, providing concrete examples of where AI is and isn’t being adopt |
| Novelty & Non-Duplication | Grok | held. Core certification-wall thesis restates existing explainers (especially the linked bioengineer.org piece) and merely drapes same-week MRO announcements on a long-known capital pattern, offering timely |
| Validation | DeepSeek | cleared. The central claim that no commercial aircraft currently flies with AI in the control loop due to certification barriers is factually supported by the provided source and aligns with known industry sta |
Sources cited: 15. Validation challenges: 0. Review cost: about $0.04. Learn how these briefings are written and verified.