Agentic AI is redefining services—now, enterprises must redesign themselves

The HFS Spring Summit 2025, held May 7–8 in New York City, delivered a reality check for enterprise leaders navigating the AI revolution. Under the bold banner—The Rise of Services-as-Software: The Agentification of Everything—the summit didn’t pull punches. This wasn’t another round of buzzwords and blue-sky promises. It was a wake-up call: enterprise services are not being enhanced by AI—they’re being redefined by it.

Through powerful sessions and candid conversations, the event tackled real-world challenges head-on—from outcome-based pricing to AI governance and the evolving role of human expertise—arming leaders with the insight and urgency needed to execute in this new era of agentified enterprise services.

Fireside chat with Phil Fersht and Ethan Mollick

Services-as-Software is not a buzzword—It’s a $1.5 trillion opportunity

Enterprise services are facing an existential pivot—and the market knows it. Services-as-Software (SaS), a term coined by HFS in September 2024, suggests that the future of services lies in software-defined, AI-native execution models that collapse the gap between tooling and outcomes. Phil Fersht drove this point home during the summit: We’re witnessing a fundamental transformation. Services aren’t being enhanced by software. They’re becoming software. In parallel, Saurabh Gupta outlined the emerging delivery category where software, AI, and services merge into continuous, outcome-led systems.

With an estimated $2 trillion in tech debt strangling innovation and 66% of enterprises dissatisfied with their software investments, trust in traditional tools is eroding rapidly. SaS is advancing faster than anticipated, opening lucrative opportunities for bold leaders across industries (see Exhibit 1).

Exhibit 1: HFS expects Services-as-Software will become a $1.5 trillion market by 2035, absorbing revenue from both traditional IT services and SaaS

Source: HFS Research, 2025

Execution, not exploration, will define the AI winners

While interest in Agentic AI and GenAI is high, real integration remains shallow and fragmented. Only 8% of enterprises use GenAI organization-wide, and just 10% report cross-departmental deployment (see Exhibit 2).

Exhibit 2: Most enterprises are stuck in GenAI exploration—only 8% have scaled organization-wide adoption

Source: HFS Research, 2025

This was reaffirmed by what we heard throughout the Summit: most enterprises still treat GenAI as a productivity experiment, not a foundational redesign. But Agentic AI isn’t built for incrementalism—it requires full integration into workflows, decisions, and outcomes to be truly effective.

However, as Dana Daher warned, GenAI doesn’t just automate—it simulates cognition. That’s precisely what makes it powerful and risky. She described a disturbing trend of “decision offloading,” where organizations risk not just handing over routine work, but surrendering critical thinking and judgment. Without clear guardrails and intentional use, enterprises may accelerate output while undermining engagement, originality, and accountability.

Ethan Mollick, Innovation Expert and Artificial Intelligence Thought Leader; Professor of Entrepreneurship, The Wharton School, and Summit keynote speaker, reinforced the urgency. In a world where everyone can access similar AI tools, “there is no technological advantage anymore.” What separates leaders from laggards is the speed of execution. The disruptor is no longer the one who uses AI—it’s the one who redesigns their enterprise to embed and govern it responsibly.

AI will scale only as fast as your people can evolve

The AI revolution isn’t a technical challenge—it’s a human one. While much of the HFS Spring Summit spotlighted agentic systems and services-as-software, one of the most urgent takeaways was more grounded: AI will only scale as fast as organizational structures and enterprise talent can adapt.

Across sessions, from Roger Lvin, Chief Executive Officer, Hitachi Digital Services, discussing Hitachi’s talent struggles to Melissa O’Brien’s panel that explored reskilling and workforce readiness, the truth that emerged was that technology isn’t the limiting factor, capability is.

That’s not just anecdotal—it’s backed by data. New HFS data confirms that talent acquisition and skill gaps are the top people-related barriers to growth
(see Exhibit 3).

Exhibit 3: Talent acquisition and skill gaps are the most significant people-related constraints on enterprise growth

Source: HFS Research, 2025, 130 Survey Participants

This shortage of human talent isn’t just slowing projects—it’s stalling transformation. Delayed timelines, increased costs, and resistance to adopting new technology are all symptoms of a workforce that’s not evolving fast enough to support the AI agenda.

Enterprise leaders must reset their mindset. The new reality is this:

• Resilience is human-first: Tech investments mean nothing if the workforce doesn’t use, supervise, and evolve them.
• Talent strategy must become AI strategy: That means shifting from workforce cost optimization to workforce capability acceleration.
• Cross-functional thinking is the new differentiator: Scaling AI and especially agentic AI requires not just engineering, but orchestration; not just skills, but contextual intelligence.

Enterprises serious about AI must design systems that elevate human expertise, not sideline it. The winners won’t be those who automate the most—they’ll be the ones who reimagine the human role at the heart of the agentic enterprise.

Enterprise execution must catch up with AI ambition

Across the HFS Spring Summit, the true missing link that emerged was that enterprises have no shortage of AI ideas, strategies, or pilots, but they are struggling to operationalize them at scale. The tech is ready, but what’s missing is the operational discipline, structural alignment, and enterprise mindset to make it work at scale.

Agentic AI is not a bolt-on—it’s a re-architecture of business and technology foundations. It collapses workflows, rewrites accountability, and demands that enterprises shift from executing tasks to orchestrating outcomes. This requires not just new platforms but also new roles, new governance, and a new enterprise mindset around value creation. Only those who execute with intent, scale with discipline, and govern with clarity will thrive.

The hard truth is that agentic AI will not fix what’s broken; it will expose it. Enterprises still clinging to outdated structures, systems, strategies, and delivery mindsets risk being left behind by those who act with clarity and intent. The time for AI experiments has passed. The time for full-spectrum redesign has arrived.

Here is what must be done:

• Stabilize the core: Clean up fragmented data, modernize systems and data foundations, and eliminate technical debt to create the foundation AI needs to scale.
• Design for orchestration, not coordination: Re-architect workflows and roles to support real-time agent interaction—collapse approval layers and empower AI-led execution.
• Define value and align delivery to it: Look beyond the safe havens of time-and-material and fixed-price models. Shift to outcome-based pricing and performance metrics that reflect real business impact. But before that, define what value means within processes and lines of business, and align outcomes to it.
• Lead from within: Become your own transformational case study. Stop waiting for vendors to come to you with ideas. Empower employees and nurture an entrepreneurial spirit around AI use—scale successful use cases internally and prove what works.
• Govern with intent, not just compliance: Build oversight models that supervise AI behavior dynamically. Focus on intent, adaptability, and trust, not just enforcing rules.

The Bottom Line: Enterprise leaders can no longer treat GenAI and Agentic AI as side projects or speculative bets. They are now the core operating paradigm.

The HFS Spring Summit 2025 wasn’t just a gathering but a strategic reckoning. Across two days of unfiltered conversations, live data, and bold insights, HFS and its attendees uncovered the truth. Services are not just being transformed by software—they are becoming software. AI is not meant to replace talent; talent must evolve to partner with it. Delivery models built on labor, legacy tools, and fixed outputs are being dismantled in favor of fluid, AI-native, outcome-driven ecosystems.

The winners will not be those who experiment the most—they’ll be those who commit the deepest to execution. The agentification era is here. The next move is yours.