Quantum Sentinel Framework: Machine Learning-Based Detection and Mitigation of Quantum-Vulnerable Cryptography

Quantum computing is a looming threat to the very security of widely recognized cryptographic techniques. More specifically, its algorithms similar to Shor will disable both RSA and ECC cryptography while the effect of security provided by symmetric methods is going to be halved. Thus, it is high time that post-quantum era preparation is done in a concerted and measured manner. The paper claims the introduction of Quantum Sentinel, which is a continuous detection, assessment, simulation, and mitigation system of quantum-vulnerable cryptography in the enterprise settings powered by AI. Quantum Sentinel makes quantum readiness practical by creating a cryptographic inventory using Certificate Transparency (CT) and optional enterprise telemetry as its sources of evidence. It then adds algorithm family, key size, certificate authority (CA) issuer, signature hash, validity window, and governance context (exposure, criticality, and data lifetime) to each endpoint. It also assesses a scenario-aware Quantum Preparedness Rating (QPR) under three adversary tiers (NEAR, MID, QDAY) to measure readiness on a unified scale. On the basis of a CT-derived TLS dataset from Saudi Arabia containing 1,241 internet-facing endpoints, the public-key ecosystem observed continues to be overwhelmingly occupied by quantum-vulnerable primitives (RSA 74.4%, ECC 25.6%). In addition, the distribution of certificate issuance is very much concentrated with a few providers (Let’s Encrypt 46.17%, DigiCert 17.49%, Sectigo 13.05%, Google Trust Services 8.62%). The resulting averages of organizational preparedness are low and worsened with stronger adversary assumptions (Mean QPR_NEAR 33.63, QPR_MID 18.55, QPR_QDAY 13.63). An AI threat-intelligence pipeline (Logistic Regression and Random Forest) can identify high-risk assets with high accuracy (Random Forest: Accuracy 0.9946, F1 0.9958, ROC-AUC 1.0), and the explainability aspect of it points to key size and algorithm as the most important features.