IT's billing model being disrupted by agentic AI

Agentic AI Tools Accelerate Billing Rate Pressure on IT Services Firms

The rapid rise of agentic artificial intelligence (AI) platforms is intensifying pressure on traditional seat-based billing models followed by global IT services firms, industry experts have said. The launch of products such as Claude Cowork and OpenAI’s Agent platform is expected to further disrupt pricing structures across the sector.

According to data from Everest Group, blended billing rates across global IT services contracts have steadily declined over the past 24 months in both the US and India. Roles impacted include web and mobile developers, full-stack engineers, remote desktop support staff, and cyber defence professionals.

For intermediate-level engineers with 4–6 years of experience, US billing rates for web and mobile developers fell to $75–91 per hour in H2 2025, compared with $77–94 per hour in H2 2024. In India, rates declined to $21–29 per hour from $22–29 during the same period.

Even premium technology consulting services, which have traditionally remained insulated from pricing pressures, have seen rate softening. US consulting billing ranges slipped to $138–181 per hour from $141–182 year-on-year.

While the rate card decline appears modest on paper, analysts warn that the real pressure lies elsewhere. Clients are increasingly demanding significant productivity-linked discounts, often requiring vendors to reinvest savings into expanding project scope or building new AI use cases. As a result, the total contract value (TCV) of a typical five-year deal shrank by 20%–35% in 2025 and is projected to decline further by 32%–44% this year.

Industry trackers HfS Research and Everest Group noted that the traditional RFP-based bidding model is losing relevance, as vendors struggle to compete in what many describe as a “race to the bottom.”

Jimit Arora, chief executive of Everest Group, said the headline rate decline masks deeper structural changes. “You will see the rates have a modest decline, but the bigger issue is how for like-for-like scope the TCV is declining. It is 30–70% productivity on baseline fees which tend to not be rate card based, but managed services constructs,” he said.

Phil Fersht, chief executive of HfS Research, highlighted that the larger shift is not just lower billing rates but fewer billable hours per unit of work as AI reduces human effort before pricing models fully reset.

“In the near term, this is negative for revenue growth because effort-based volume no longer scales the way it once did,” Fersht said. “The impact on margins is more nuanced. Firms that simply absorb AI into delivery without changing pricing will see margin pressure. Those that redesign delivery models, structurally reduce effort, and shift toward outcome-linked or productised services can protect, and in some cases expand, margins.”

Pareekh Jain, founder and chief executive of IT research firm EIIR Trend, said traditional fixed-price contracts and time-and-material rate cuts are ill-suited for the uncertainty introduced by AI-led productivity gains. He suggested that usage- or subscription-based pricing models could offer a more sustainable alternative.

Experts broadly agree that the sector is undergoing a fundamental business model shift — from people-intensive, effort-based projects to AI-enabled, outcome-driven service delivery models — reshaping how IT services are designed, delivered and monetised.