IT Services in the Age of Agentic AI: Underwriting Through a Structural Shift

• June 22, 2026
• North America
• Technology, Media, and Telecom

• Share This Page
• LinkedIn
• Twitter
• Facebook

June 22, 2026 - Few sectors are generating more contradictory signals than IT services right now. On one side, agentic AI is being framed as an existential threat to a roughly $1.9 trillion industry built on labor arbitrage and time-and-materials billing.¹ On the other, private capital continues to flow into managed service provider (MSP) roll-ups, with a recent wave of recapitalizations alongside evergreen and AI-affiliated capital entering the space. The two pictures are less contradictory than they appear, but the contrast is real: AI is reshaping the economics of services delivery in ways the market is still absorbing, while capital keeps underwriting the long-term value of well-positioned operators. For lenders, the more useful question is not which signal wins, but how to think about businesses under the IT services umbrella when the answer is genuinely uncertain. This is a sector WhiteHorse has been active in for years, and one where we continue to see meaningful deal flow.

The headline data still favors consolidation. Even as the disruption narrative builds, IT services remains the single largest category of worldwide IT spending, larger than software, devices, or communications, and the segment is still growing.¹ For the right kind of operator, the demand backdrop can be healthy.

What is less clear is how the cost structure and value proposition evolve from here. IDC expects 30% of all IT services contracts to be outcome-based by 2029, and 30% of IT services to be delivered as modular, platform-enabled products rather than billable hours.² Gartner frames the shift similarly, from selling tools and labor toward selling outcomes priced on value rather than cost-plus margin on hours.³ Forrester is consistent: AI is unlikely to kill IT services, and the winners will be the firms that build proprietary AI platforms and deepen domain expertise, not those competing on commodity labor.⁴ From what we have seen, MSPs built on offshore body shops, basic helpdesk, and undifferentiated implementation work appear to face real margin compression, while those close to mission-critical workflows, delivering measurable outcomes, seem to be well positioned to absorb AI as a tailwind. How durable each dynamic proves to be is still an open question.

A Bifurcated Services Market

In our view, "IT services" is best understood as a label covering several distinct businesses, and within any one business, individual segments often carry very different AI exposure. The second point matters as much as the first. A single borrower may run a multi-year managed services contract that AI strengthens, a T&M staff augmentation book that AI compresses, an offshore helpdesk operation that AI partially displaces, and a cybersecurity consulting practice that AI accelerates, all at once. Treating that company as one risk profile misses the picture.

Looking at revenue mix at the segment level, the patterns are roughly these. Recurring managed services under multi-year master service agreements, particularly those anchored in cybersecurity, compliance, or other mission-critical workflows, tend to look more durable. AI here shows up mostly as a productivity lever, helping the operator serve more endpoints per technician and supporting margins. A simple example: an MSP managing endpoint detection and response for a regional healthcare customer is unlikely to be displaced by an agent that drafts incident summaries faster. More often, that MSP deploys the agent itself to handle more clients with the same headcount.

A related tailwind sits on this end of the market. A 2025 Gartner survey found only 14% of enterprises plan to handle their AI work primarily with internal resources, while a majority intend to rely on service providers for at least half of it.³ Smaller, less sophisticated buyers especially lean on outside help to clean data, deploy tooling, and adopt AI at all, which is real incremental demand. None of this makes the segment risk-free: where services are priced per seat or per device, the same efficiency that lifts margins can erode pricing power if customers come to expect the savings.

Project-based and T&M work spans a wider range. Complex advisory, cyber consulting, and digital transformation engagements often retain pricing power, because the judgment and risk-bearing component is hard to automate. Undifferentiated implementation and staff augmentation look more exposed, particularly as offshore delivery models shift away from labor arbitrage toward higher-skilled work, and offshore helpdesk and basic ticketing sit at the most exposed end. Many providers in these categories are themselves building AI accelerators and productized tools that, by design, compress their own billable hours. It is the right strategic response, but it complicates the picture: short-term revenue may be cannibalized for longer-term positioning, and we watch closely how operators are managing that transition.

We have also seen vertically focused firms, particularly those serving highly regulated end markets such as healthcare, financial services, legal, and defense, hold up well across macro and technology cycles. Regulatory complexity, audit cycles, and the cost of getting it wrong create switching costs largely independent of what AI does to delivery economics. A defense-focused MSP managing classified or controlled-unclassified-information environments, for example, is solving a problem AI tooling does not eliminate on its own. If anything, AI introduces new compliance and auditability questions those providers have to navigate.

One related dynamic worth flagging: the broader shift toward outcome-based pricing. By our read, most IT services businesses are in the early innings of this transition. There is not yet a settled playbook for how the pivot is sequenced, how risk is priced, or what the steady-state economics look like. The market is still calibrating to a structural shift, and we expect that calibration to take time.

How We Are Thinking About Underwriting

Against this backdrop, quality of management matters more than ever. The pace of change in delivery models, pricing structures, and AI tooling is widening the gap between operators actively reshaping their business and those who are not. We place a premium on teams that are clear-eyed about which parts of their revenue base are most exposed, willing to invest ahead of the curve, and unafraid to evolve pricing and delivery. WhiteHorse's relationship-driven model, partnering closely with both sponsor-backed and founder-owned operators, has historically lent itself well to backing teams like that.

AI has also reshaped how we underwrite. We have dedicated AI resources across our broader platform and continue to add experienced AI talent, all of which we put to work in our own diligence: pressure-testing revenue mix at the segment level, stress-testing management's view of which segments are most exposed, and synthesizing primary research far faster than was possible even a year ago.

When we look at IT services credits, we think about:

• Revenue Mix at the Segment Level: What portion of revenue sits in long-dated contractual managed services, recurring transactional work, and project/T&M, and how each is likely to behave under different AI scenarios.
• Customer and Workflow Stickiness: The role the borrower plays in the customer's operations, the cost and risk of switching, and the presence of regulatory, compliance, or cyber anchors that reinforce retention.
• Management's Posture Toward Change: How the team is thinking about pricing model evolution, AI in delivery, talent mix, and where they are investing.
• Vertical Focus and End-Market Characteristics: Whether the business has built defensible positioning in specific verticals, particularly regulated ones, versus competing horizontally on price.

None of these are gating criteria. A business with more AI-exposed revenue characteristics is not necessarily, in our view, uninvestable. It may simply warrant lower leverage, tighter structure, or different pricing.

Comfortable With the Question Marks

It is tempting in moments like this to take a strong view on where this all lands, to argue either that AI hollows out IT services or that consolidation continues to compound regardless. We do not pretend to know exactly how this plays out, and we are skeptical of anyone who says they do. What we do believe is that demand for the right kind of IT services business is, if anything, strengthening. Who emerges as winners and losers, and how delivery and pricing models ultimately settle, will become clearer over time. In the meantime, the businesses likely to do well through whichever scenario unfolds are the ones with sticky customer relationships in markets where the cost of failure is high, run by management teams willing to evolve the model as the ground shifts. Across both sponsor-backed platforms and founder-owned operators, those are the businesses, and the partnerships, we keep looking for.

Citations

1. Gartner, "Gartner Forecasts Worldwide IT Spending to Grow 13.5% in 2026, Totaling $6.31 Trillion" (April 2026); IT services is the largest spending category at approximately $1.87 trillion. https://www.gartner.com/en/newsroom/press-releases/2026-04-22-gartner-forecasts-worldwide-it-spending-to-grow-13-point-5-percent-in-2026-totaling-6-point-31-trillion-dollars
2. IDC, "Charting the Agentic Future of IT Services: From Projects to Platforms," FutureScape: Worldwide Services 2026 (December 2025). https://www.idc.com/resource-center/blog/charting-the-agentic-future-of-it-services-from-projects-to-platforms/
3. Gartner, "Impact of AI on Services" (March 2026). Enterprise resourcing data from the 2025 Gartner Digital Services Buyer Behavior Survey, as presented therein.
4. Ashutosh Sharma, "The 'Death' of IT Services Is Greatly Exaggerated," Forrester (February 24, 2026). https://www.forrester.com/blogs/news-of-it-services-death-is-greatly-exaggerated/

Note: The views expressed herein are those of the author as of the date of publication, are subject to change without notice, and do not necessarily represent the views of WhiteHorse Capital. This material is provided for informational purposes only and does not constitute investment advice or an offer or solicitation to buy or sell any security. Certain information has been obtained from third-party sources believed to be reliable, but WhiteHorse Capital makes no representation as to its accuracy or completeness. This material may not be reproduced or redistributed without the prior written consent of WhiteHorse Capital.