Secure AI-SDLC for Critical Infrastructure: Operationalizing the NIST AI RMF with Evidence-Driven Controls
Bibliographic record
Abstract
Critical infrastructure systems increasingly rely on artificial intelligence for decision-making, yet current development practices inadequately address security risks unique to AI systems. This paper presents Secure AI-SDLC, a comprehensive security framework that operationalizes the NIST AI Risk Management Framework with structured software development lifecycle controls for cyber-physical and safety-critical environments. Our approach maps AI-specific threats to actionable engineering controls and required evidence artifacts across five lifecycle phases: govern, design, build, test, and deploy/monitor, aligned with ISO/IEC 42001, IEC 62443, and NERC CIP. Validation through a 16-week deployment on an industrial monitoring pipeline in the energy sector demonstrated measurable improvements, including a 12.3% increase in adversarial robustness and a 40.2% reduction in mean time to detect anomalies. Mean time to recovery decreased from 4.2 hours to 1.8 hours. The proposed framework addresses the gap between high-level AI governance and practical implementation by providing concrete, auditable mechanisms for building and operating trustworthy AI systems in high-consequence environments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.017 | 0.028 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.001 |
| Science and technology studies | 0.001 | 0.006 |
| Scholarly communication | 0.005 | 0.008 |
| Open science | 0.003 | 0.007 |
| Research integrity | 0.002 | 0.004 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".