Abstract : Artificial intelligence (AI) is rapidly transforming how products are conceived, built, and deployed - but the traditional product development lifecycle (PDLC) was not designed for probabilistic, data-dependent, and continuously evolving systems. Existing responsible AI frameworks - including the NIST AI Risk Management Framework (AI RMF), the EU AI Act, and global ethics guideline surveys - establish important governance principles but do not specify how product managers should operationalize these principles across the PDLC phases of problem framing, requirements definition, data strategy, design, development, validation, launch, and post-deployment monitoring. This paper proposes the Responsible AI Product Management (RAIPM) framework, which positions ethics, compliance, and trust as integrated product management disciplines rather than downstream compliance activities. RAIPM introduces five operationalizable mechanisms: trust-aware requirements specification, compliance-by-design governance architecture, documentation-centered governance artifacts, outcome-based accountability models, and bias-integrated validation protocols. The framework is empirically evaluated through a 16-week enterprise AI product governance program spanning seven AI product initiatives, demonstrating a 46% improvement in ethical compliance score, an 83% bias audit pass rate compared to 51% at baseline, a 94% governance artifact completeness rate, and a 31% improvement in user trust indicator. These results establish RAIPM as a practically validated, governance-aligned framework for product managers operating AI products under emerging regulatory obligations and stakeholder accountability expectations.