McKinsey’s 2025 AI Report confirms what we’re seeing in the market: enterprise AI has bifurcated. A small elite, roughly 6% of companies, are capturing transformative value (5%+ EBIT impact), while the remaining 94% are trapped in pilot (pre-production) purgatory. The gap isn’t closing. It’s widening.

The numbers tell the story. While 88% of companies use AI in at least one function, only one-third have scaled beyond pilots. Most can’t quantify impact: just 39% report any EBIT effect, and most of those claim less than 5%. Meanwhile, high performers are pulling away. They invest over 20% of digital budgets in AI, deploy across multiple functions, and fundamentally redesign workflows rather than bolting AI onto broken processes. When executives personally own AI strategy, profitable deployments are three times more likely. But most organizations lack both the leadership engagement and operational discipline to follow.

AI agents show the same pattern. While 62% experiment with agents, only 23% are scaling them, and most scale in just one or two functions. Adoption is concentrating in IT and knowledge management, service desks and research workflows, where use cases have matured fastest. Everywhere else, agents remain speculative.

Risk is no longer theoretical. Half of organizations have experienced negative consequences from AI, with inaccuracy leading the list. High performers report more failures, particularly around IP infringement and regulatory compliance, because they’ve deployed more aggressively. The companies moving fastest are discovering that guardrails, validation processes, and compliance frameworks aren’t optional—they’re bottlenecks to scale.

Workforce expectations are all over the map. Thirty-two percent of companies anticipate headcount reductions of 3% or more in the next year. Thirteen percent expect increases. Forty-three percent see no change. Yet nearly everyone is hiring software and data engineers to build AI systems, creating a talent crunch that slows deployment even as companies plan for automation-driven cuts.

Company size determines outcomes. Nearly half of $5B+ companies have reached the scaling phase. Among sub-$100M companies, only 29% have. Large enterprises have resources but struggle with organizational inertia. Smaller companies have agility but lack capital and specialized talent. Both are stuck, just differently.

The opportunity is clear: the 94% need a different path forward.

High performers redesign workflows, secure executive ownership, mitigate risk systematically, and measure relentlessly. Most companies can’t replicate this playbook because they lack the talent, the budget, or the organizational will. The next wave of value creation belongs to companies that make transformation accessible without requiring armies of AI engineers or multiyear reorganizations.

Three bets stand out:

Vertical transformation platforms.

Horizontal AI tools assume customers will figure out deployment. They won’t. The winners will own specific workflows end-to-end (service desk automation, clinical documentation, financial underwriting, legal research) delivering repeatable, high-margin services that drive measurable EBIT without requiring customers to become AI experts. Success requires deep domain knowledge, not better models.

1. Vertical Transformation Platforms: The Revenue & Efficiency Bet

Executive Rationale: This bet is about moving from fragmented, costly operations to high-margin, predictable service delivery that drives immediate EBIT impact.

Risk and governance infrastructure.

As companies scale, inaccuracy, compliance failures, and IP infringement move from edge cases to business-critical risks. High performers experience more failures because they deploy more use cases. Demand for validation frameworks, automated guardrails, compliance layers, and output warranties will accelerate. The companies that can credibly de-risk AI deployment, especially in regulated industries, will unlock adoption that’s currently stalled.

2. Risk and Governance Infrastructure: The Trust & Compliance Bet

Executive Rationale: This bet transforms AI from a potential liability (hallucination, inaccuracy, compliance failures) into a strategic asset, protecting brand trust and avoiding massive financial and regulatory penalties.

Talent leverage platforms.

The AI talent bottleneck is real and worsening. Companies need scarce engineers to build systems that may ultimately reduce headcount elsewhere. Low-code agent builders, automated workflow redesign tools, and self-service AI infrastructure that reduce dependency on specialized talent will determine which companies can scale and which stay trapped in pilots. Democratizing what currently requires PhDs isn’t a nice-to-have. It’s the unlock.

3. Talent Leverage Platforms: The Velocity & Scalability Bet

Executive Rationale: This bet solves the strategic bottleneck of the AI talent crunch, ensuring the organization can rapidly deploy and scale AI initiatives without hiring thousands of scarce, expensive PhDs or disrupting the existing workforce.

The conclusion is clear:

For you to move your contact center and CX operations out of “pilot purgatory” and into the “elite 6%,” you must stop experimenting with general AI tools and start adopting packaged, verticalized solutions that redesign your core workflows, come with built-in risk governance, and are easy for your existing operations teams to manage and scale.

The bifurcation won’t reverse on its own. The 6% who’ve figured it out will keep pulling away. The 94% who haven’t need solutions that work without transformation theater, meaning packaged offerings that deliver results, not frameworks that promise them. The market is massive, the need is urgent, the solves are complex, but the window is open. The companies that solve for workflow transformation, risk mitigation, and talent leverage simultaneously won’t just capture value. They’ll define the next decade of enterprise AI.