Well integrity decisions in mature offshore fields increasingly rely on quantifying how corrosion-driven wall thickness loss in tubulars translates into time-varying failure risk when inspection data are scarce. This work proposes a data-updated, time-dependent reliability framework that links corrosion degradation to API TR 5C3 limit states for failure under internal and external pressure [1]. The degradation process is represented as a Wiener degradation process and benchmarked against the NORSOK M-506 CO2 corrosion-rate model [2]. A single corrosion log provides thousands of wall-thickness measurements along the well depth at one inspection time. However, logging operations are costly and only a limited number of wells are logged, typically only once during their entire operational life. Under a constant drift assumption across depth, Bayesian inference updates the Wiener drift, diffusion and measurement uncertainty. Posterior sampling is performed via adaptive Bayesian Updating with Subset Simulation [3], which avoids prescribing a likelihood-scaling constant and reduces model evaluations through intermediate conditional levels. Posterior predictive thickness trajectories are propagated using a monthly first-passage Monte Carlo scheme over a 240-month horizon. Non-failed realizations are treated as right-censored, enabling nonparametric estimation of failure probability, hazard and remaining useful life with uncertainty bands, prior-versus-posterior comparisons and quantified information gain from inspection updating. The approach is demonstrated for an offshore injection well in the Campos Basin, eastern Brazil. Conditioning on the measured data yields an immediate reduction in near-term conditional hazard and a marked tightening of predictive uncertainty, followed by renewed hazard growth as degradation progresses. In the case study, comparison with inspection logging data showed that the NORSOK-based model leads to conservative reliability estimates. Additional inspection logs can be assimilated to refresh the posterior of degradation parameters and to update component-level reliability parameters. Overall, the workflow links inspection evidence to time-dependent reliability, supporting risk-informed inspection timing and life extension screening.