AI in Action Beyond Experimentation to Transform Industry 2025

Defining depth of AI adoption TABLE 1 AI organizational adoption phases Description Metrics/indicators Examples Phase 1 Initial and ad hocOrganizations are just beginning their AI journey, hindered by barriers such as regulatory constraints, organizational risk aversion and/or a lack of awareness of AI’s potential. –Data quality and accessibility –Basic data and compute infrastructure –Limited to AI sandbox experiments Phase 2 “Thousand flowers bloom”Organizations are running multiple AI experiments, often disconnected from core business strategy and possibly driven by individual teams or tech-savvy leaders. –AI project volume –Increased investment in AI talent –Workforce training and upskilling efforts –Initial governance practices –Use cases prioritized, with multiple use cases piloted through MVP (minimum viable product) development and production and some identified for scaling Phase 3 End-to-end reinventionOrganizations are moving beyond experimentation and beginning to see measurable value from deploying AI at scale within a specific business domain. –Presence of a formal AI strategy –Robust data governance –AI-enabled process improvements –Measuring ROI of AI projects –Reinvention of marketing strategies, supply chain and customer service –Functional cognitive brain Phase 4 Enterprise-level reinventionOrganizations are aligning AI initiatives across multiple functions, supported by foundational infrastructure, robust data governance and workforce upskilling, to ensure effective integration of AI across business units. –Established ethical AI board –Significant customer impact –Continuous improvement culture –Positive impact on business outcomes –Data products consumable across functions –Enterprise cognitive brain –Functional and governance silos broken down Phase 5 Value chain reinvention (future vision)AI initiatives extend across the entire value chain, creating innovative collaborations with partners, suppliers and even competitors. –AI as a core business enabler –Continuous improvement culture –Significant stakeholder impact –Substantial impact on business outcomes –Continuous monitoring and reassessment –Strategic outcome evaluation Source: Accenture analysis in collaboration with World Economic Forum consumer industries team. As AI moves beyond experimentation, a broader shift towards enterprise-level reinvention of business and operating models will emerge. Early signs of transformation are already visible, with leading organizations implementing changes to transition towards AI-enabled models, and AI-native models are even emerging. New intermediaries and disintermediation powered by AI could disrupt incumbents while driving shifts in value dynamics and the development of emerging business ecosystems. Organizations should prioritize applications that deliver tangible, measurable value and focus on scaling their adoption across the business. This requires identifying high-impact use cases, optimizing them, building a strategic roadmap for broader implementation and, importantly, learning from industry peers and sharing best practices to accelerate growth, avoid common pitfalls and unlock AI’s full potential. Semiconductor firm AMD, in partnership with a major software company, enhanced its sales- order capabilities with a genAI supply chain troubleshooter tool that offers detailed insights into order commitments, product allocations and supply issues. The tool analyses over 10,400 orders annually, saving AMD some 3,120 hours in productivity and cutting the time and cost of its root-cause analysis by 90%.29 CASE STUDY 9 Emerging transformation AI Governance Alliance: Transformation of Industries in the Age of AI 14