BGP, the Internet’s inter-domain routing protocol, is vulnerable to prefix hijacks due to tamperable prefix–origin bindings. RPKI addresses this by cryptographically binding prefixes to authorized ASes, enabling Route Origin Validation (ROV). Given RPKI’s critical role in Internet security, identifying the proportion of ASes that deploy ROV is a key research question. However, measuring ROV deployment is difficult due to limited visibility into private AS configurations. Existing methods suffer from low accuracy, restricted coverage, and the inability to detect hidden ASes whose behavior is masked by upstream ROV deployment. To address these limitations, we propose RIFT, a novel inference method based on temporal fingerprinting. RIFT leverages the insight that periodic ROA retrieval creates distinctive temporal patterns in the routing behavior of ROV-deployed ASes. Experiments show that RIFT achieves 94% accuracy and an F1 score of 0.88 in identifying ROV deployment.