Don’t wait for 2030: Reinvent your workforce or lose the AI race

The IT services workforce is about to flatline. By 2030, total headcount across enterprises, service providers, and GCCs will stop growing as AI absorbs legacy debt work.

This is a system-wide call to action for the entire IT services value chain to redesign work before AI rewrites it for them.

The Zero-Addition Era is coming—and every actor in the value chain has 24–36 months to redesign how work is created, delivered, and governed before automation renders current operating models obsolete. HFS analysis shows what survival in this reset demands and what every player must do now to stay relevant.

The rapid infusion of AI into IT services delivery has triggered one of the most consequential debates of our time: What happens to the workforce when technology begins to do the work? Recent layoffs by major providers have fueled speculation of an industry-wide contraction.

The data, however, tells a more measured story. HFS analysis shows that while an “AI job collapse” is not imminent, the window to prepare for structural change is closing fast. By 2030, the industry will enter a Zero-Addition Era, where total headcount remains broadly stable even as the nature of work, roles, and value creation transform entirely. To thrive in this era, enterprises, GCCs, and service providers must rebuild their talent model by redefining roles, re-engineering learning, and rethinking how performance and value are measured. Those that fail to reform will hit a productivity cliff, while AI-native competitors double output without adding headcount.

You have 36 months to reinvent your workforce before growth flatlines

The industry stands at the edge of a structural realignment that will unfold across three phases: from debt-driven growth in the near term to automation-led recomposition, culminating in an intelligent equilibrium where human capability and machine intelligence operate as one
(Exhibit 1).

Over the next two to three years, all industry cohorts will be compelled to redesign how work is created, delivered, and governed across the value chain.

Exhibit 1: Legacy debt is driving headcount growth that AI will soon erase

Source: HFS Research analysis
Baseline 2025 workforce of ~15 million covers IT services providers, global capability centers (GCCs), and retained enterprise IT.
Estimates are derived from publicly reported headcount data and HFS econometric modeling of workforce trajectories across internal and external delivery environments.
The automation impact curve models real-world adoption patterns observed in enterprise operations—both in-sourced and outsourced—calibrated against current productivity gains and talent substitution rates, combined with an estimated 6% annual organic demand growth.
The post–AI-era uplift reflects HFS scenario modeling that assumes roughly one new role created for every five roles automated, concentrated in orchestration, assurance, and AI-enabled delivery design.

Phase 1: Debt-driven growth (2025–2028)

Legacy debt is the only thing keeping your headcount growing.

Over this period, employment across IT services will grow modestly, as it has over the past few years. However, it will be driven less by new digital demand than by the accumulated weight of legacy technology debt. Outdated applications, fragmented architectures, and manual dependencies will continue to require human oversight, creating a paradoxical cushion: While automation expands, people remain essential to maintaining the stability of systems.

AI will be ubiquitous, but its impact will be uneven. Generative tools and AIOps platforms will accelerate testing and monitoring, yet most enterprises will still operate hybrid environments with new intelligent systems layered on top of legacy stacks. Automation may even increase complexity, adding new monitoring and validation demands.

This period marks the final phase of volume-driven growth. Orchestration roles will emerge to design and oversee automated processes rather than execute them manually. By 2028, as AI-driven operations mature, the balance will shift toward large-scale recomposition.

Phase 2: The recomposition plateau (2028–2030)

By 2028, the L1-L4 pyramid will collapse—are you ready?

By 2028, the debt that once sustained headcount will have shifted from people to technology. Agentic and generative AI systems will now manage much of this “debt layer” directly by autonomously monitoring systems, executing workflows, predicting failures, and remediating incidents. As AI assumes this burden, the long-standing correlation between effort and employment will break.

The traditional L1–L4 pyramid, organized by a hierarchy of labor, will start to widely collapse into a new delivery fabric organized across two axes (shown in Exhibit 2).

Exhibit 2: From pyramids to lattices—the IT services industry needs to brace for structural workforce recomposition

Source: HFS Research analysis, 2025

The traditional hierarchy of tasks and titles is giving way to a lattice of intent and fluency, where value is created not by role level but by how effectively humans and AI collaborate to achieve outcomes. Every IT leader must understand how this emerging delivery fabric and the two new defining axes will recompose their workforce for the AI era.

Axis 1: Work intent (the “what” of work)

Axis 1 defines why the work exists and what outcome it achieves.

• Create: Designing and building new AI-native systems, platforms, and services that blend digital, physical, and data-driven environments.
• Change: Continuously adapting and improving existing systems through migrations, stress-testing, reconfiguration, and optimization.
• Operate: Supervising AI-enabled operations, self-healing infrastructure, and closed-loop service ecosystems.
• Assure: Governing performance, compliance, and business value while ensuring ethical, resilient, and trustworthy operations.

Axis 2: AI fluency (the “how” of work)

Axis 2 defines how effectively humans and AI collaborate to deliver that outcome.

• Practitioner: Applies AI tools and copilots to enhance productivity within defined workflows.
• Orchestrator: Designs and manages workflows where AI and humans share tasks.
• Architect: Builds scalable AI-native systems and pipelines that embed intelligence into delivery.
• Governor: Oversees performance, trust, and responsible value creation across intelligent ecosystems.

Structurally, the workforce now resembles a diamond with a smaller base of AI-fluent practitioners, a broad middle of orchestrators and composers, and a focused apex of architects and governors aligning intelligent delivery with business outcomes.

Early signals of this recomposition are already visible:

• Operate–Practitioners: AIOps and reliability engineers who supervise automated loops and manage exceptions.
• Change–Orchestrators: Forward-deployed engineers who design, customize, and embed AI systems directly in client environments, acting as the connective tissue between product intelligence and delivery outcomes.
• Create–Architects: Platform and LLM engineers building scalable AI-native architectures and autonomous service layers.
• Assure–Governors: Value-assurance and responsible-AI leaders overseeing performance, trust, and compliance.

Phase 3: Intelligent equilibrium (2030+)

Headcount will stabilize, and the lattice of intent and fluency will take full shape.

By the early 2030s, a fully synergized human-machine model will be in place. The workforce will stabilize in size but reorganize as a lattice of work intent × AI fluency, the structure that underpins intelligent operations (see Exhibit 3).

Exhibit 3: The lattice of intent and fluency: what the AI-age workforce will look like

Source: HFS Research, 2025

Humans and AI will operate as one delivery fabric where machines execute and adapt, while people design, govern, and assure. This will be the intelligent equilibrium, a state of headcount stability but constant reinvention, where competitiveness depends on how human and machine capabilities amplify one another.

The industry must use the next 2–3 years to recode its workforce for the AI economy

AI will redefine delivery models, but people will continue to define organizational identity. Firms that use the short window of the next two to three years to re-architect talent, and not just technology, will own the next decade.

Stop cutting graduate hires; you’re destroying your future talent pool

Entry-level hiring must evolve, not erode. Reducing graduate intakes may protect margins in the short term, but it weakens the industry’s long-term capacity for renewal and adaptability. The goal is not to eliminate entry-level roles, but to redefine them for an AI-shaped workplace by incorporating technical literacy, AI fluency, and systems thinking from the start.

Key pathways for entry-level hires include

• Apprenticeship programs blending code, data, and orchestration skills.
• Automation academies to redeploy L1 operators into AI-composer roles within 18 to 24 months, equipping them to design and manage agentic workflows.
• Mid-career reskilling, as the 3- to 7-year talent band becomes the core of orchestration and assurance.
• Governance guilds uniting technology, risk, and finance professionals to embed value assurance across AI-enabled delivery.

The L1-L4 ladder is dead; organize around work intent, not titles

The traditional L1–L4 job ladder will dissolve into intent-based and fluency-driven roles. Every role must connect to one of four work intents (create, change, operate, assure) with explicit expectations of AI fluency. This is not a replacement but a repurposing of talent, evolving human contribution from repetition and execution to reasoning and orchestration.

Building a unified role taxonomy that spans engineering, operations, and governance will bridge the long-standing divide between “delivery” and “support.”

Certifications are worthless; build continuous, experiential learning ecosystems

As roles evolve from a hierarchical structure to one driven by intent, learning must evolve from training to transformation. The next phase of workforce competitiveness will hinge on fluency—the ability to design, govern, and collaborate with intelligent systems while exercising judgment, empathy, and adaptability.

The future workforce must combine technical and analytical depth with human dexterity: listening, accountability, ethical reasoning, and calm decision-making under uncertainty. Transformation will depend as much on how leaders guide people through change as on how fast they deploy AI.

To enable this shift, firms must create adaptive learning ecosystems that allow people to move fluidly across work intents:

• Create: AI-native design, data modeling, creative experimentation.
• Change: Process re-engineering, low-code orchestration, stakeholder collaboration.
• Operate: AI-operations supervision, exception handling, situational decision-making.
• Assure: Responsible-AI governance, compliance analytics, value measurement.

To make this transformation a reality, every enterprise must rewire how people build capabilities. The workforce will no longer climb a job ladder but rather move fluidly across overlapping domains of fluency that connect human skills with machine capabilities. This evolution unfolds through four overlapping capability domains, defining the new human–machine delivery fabric.

Exhibit 4: Build fluency across four capability domains to power the human–machine workforce

Source: HFS Research, 2025

Learning can no longer be measured in hours or certifications. It must be continuous and experiential, powered by generative tutors, simulation labs, and fluency badges that measure applied capability, not tenure.

Activity metrics will kill you; measure fluency and outcome density

In the Zero-Addition Era, the goal will be to scale value density and redeployment velocity of the workforce. Firms must replace activity-based metrics with fluency- and outcome-based measures that assess how effectively humans and AI work together.

Exhibit 5: Redefine performance by replacing activity metrics with fluency and outcome measures

Source: HFS Research, 2025

Service towers are dissolving; merge infrastructure, apps, data, and security now

The boundaries between technology and operations will dissolve. The traditional towers of infrastructure, applications, data, and security will evolve into unified business systems that combine orchestration and execution.

Each will reconstitute as a micro-lattice of AI-enabled capabilities:

• Infrastructure and cloud: Autonomous fabrics that self-monitor, self-heal, and self-optimize.
• Applications: Adaptive systems continuously refactored through generative and low-code automation.
• Data and AI: The connective tissue linking insight, automation, and assurance.
• Security and risk: Embedded digital-trust systems governing AI behavior and data lineage.
• Enterprise platforms: Abstraction from configuration to composition of unified business operating systems.

This shift will unfold through persona-first redesigns, shared IT–ops structures, and AI-driven workflows that bring technology and operations together around end-to-end outcomes.

The Bottom Line: The Zero-Addition Era is not a slowdown—it’s a reset.

The Zero-Addition Era starts in 2028. By then, AI will absorb your legacy debt work, and headcount growth will stop permanently. Use the next 36 months to rebuild your talent pipeline, redesign roles around AI fluency, and create continuous learning systems, or watch your workforce become irrelevant while competitors scale value density without adding people. The choice is simple: Reinvent talent now or manage decline later.