Blockchain technology has become a fundamental enabler for secure and decentralized data management in fog computing and the Internet of Things (IoT). However, conventional consensus mechanisms such as Proof-of-Work (PoW) and Proof-of-Stake (PoS) suffer from limitations including high energy consumption, centralization risks, and limited scalability, making them inefficient for dynamic, resource-constrained fog environments. To address these challenges, this research proposes a Quality of Service (QoS)-aware privacy preservation framework that integrates blockchain, Attribute-Based Encryption (ABE), and machine learning. A Grey Wolf Optimization (GWO)-enhanced hybrid consensus model is introduced, combining PoW-based computational security with PoS-driven energy efficiency, dynamically balancing workload and trust-based miner selection. The proposed GWO-powered trust evaluation optimizes node selection based on trust score, mining efficiency, and energy consumption, ensuring enhanced security against Sybil attacks and other adversarial threats. Furthermore, a Modified Attribute-Based Encryption (ABE) scheme is incorporated to provide fine-grained access control and computational efficiency for privacy-preserving data sharing in fog computing. The modified ABE integrates lightweight cryptographic operations, reducing computational overhead while ensuring secure, policy-based access control. Experimental evaluations demonstrate that the proposed framework reduces communication delay by 16.5%, improves energy efficiency by 10.4%, and increases throughput by 23.5% compared to existing state-of-the-art models such as DRLBTS, QoS_ML_DSS, and SLGAF. These results highlight the effectiveness of the model in providing a scalable, secure, and energy-efficient blockchain solution for privacy-sensitive fog computing applications in healthcare, smart cities, and industrial IoT.