As dynamic pricing becomes increasingly person- alized through machine learning, concerns about fairness have moved to the forefront of algorithmic pricing design. While traditional fairness definitions in automated decision systems have focused on group parity or individual similarity, pricing introduces domain-specific challenges such as loyalty penalties, behavioral exploitation, and price envy. This paper proposes a unified framework that quantifies and balances four fairness notions in personalized pricing: group fairness, individual con- sistency, behavioral (loyalty-based) fairness, and envy-freeness. The approach integrates these soft constraints into a multi- objective optimization model that augments the standard revenue maximization objective.

We simulate a heterogeneous population of agents with dis- tinct behavioral traits and compare four policies: unconstrained pricing, group fairness-constrained pricing, loyalty fairness, and a combined multi-fair policy. Results show that fairness-aware pricing substantially reduces loyalty gaps and envy violations while maintaining over 96% of unconstrained revenue. The multi-fair policy achieves the most balanced performance across fairness dimensions, demonstrating that equity and efficiency can be jointly realized in algorithmic pricing systems. The findings provide a theoretical and empirical basis for fairness-aware pricing with implications for SaaS, marketplaces, and digital services.