New research from KPMG and the University of Texas at Austin reveals a clear divide between businesses extracting real value from AI and those still in experimental mode. The highest-impact AI users treat artificial intelligence as a reasoning partner—not just a productivity tool—and are scaling AI agents while strategically managing multiple models to connect spending directly to business outcomes. The ability to work effectively with AI as a thought partner is a learnable skill that organizations can teach at scale. The findings signal a maturation of enterprise AI adoption, with leading organizations moving beyond automation and cost-cutting to pursue new revenue opportunities. Rather than deploying AI narrowly for efficiency gains, top performers are building AI-native workflows that fundamentally change how teams work and make decisions. Meanwhile, the AI infrastructure landscape continues to shift: Anthropic is preparing for an IPO before Thanksgiving, Meta's Muse momentum has spurred investor enthusiasm, and Google has launched its first custom AI chips into orbit as part of its broader push into space infrastructure.