Sixty-three percent of employers say skills gaps are the biggest barrier to their transformation efforts, not budget, not technology maturity, not leadership buy-in.¹ Yet most companies are still trying to solve an AI talent problem with a pre-AI hiring process: write a job description, get headcount approved, find one person, place them on an existing team.
That model isn’t broken. It’s just no longer sufficient.
At IDR, we sit inside these conversations every day, with the technology leaders deciding which roles to make permanent, which to bring on for a single project, and which they can’t yet define with enough confidence to do either. That vantage point is what this piece draws on. And it points to a clear conclusion: to realize the value of AI, organizations need a more flexible technology talent model, one that deliberately combines direct hiring, staff augmentation, contingent staffing, contract-to-hire, and specialized project support based on the work that actually needs to get done.
The case for a more flexible talent model doesn’t depend on the assumption that AI will create more technology jobs than it eliminates. It begins with a more practical observation: AI is changing where demand appears, which skills are scarce, and when organizations need them.
U.S. software development job postings increased by nearly 15 percent between February 2025 and June 2026, even as overall job postings declined 7 percent.² The rebound was concentrated in experienced and AI-related positions. Senior roles accounted for 71 percent of the increase in software development postings between May 2025 and May 2026, and positions mentioning AI in the title accounted for 37 percent, with some overlap between the two groups.²
The broader technology labor market shows the same underlying demand. The U.S. Bureau of Labor Statistics projects approximately 317,700 openings annually across computer and information technology occupations between 2024 and 2034, driven by both employment growth and the need to replace workers who permanently leave those occupations.³
Even the physical infrastructure behind AI is reshaping demand. The U.S. Department of Energy reports that data centers consumed roughly 4.4 percent of total U.S. electricity in 2023 and could account for between 6.7 and 12 percent by 2028, with data center expansion and AI application growth cited as key drivers.⁴
For CIOs, CTOs, digital transformation executives, and other leaders responsible for technology execution, the implication isn’t simply that AI will produce more or fewer jobs. It’s that the composition of the technology team is changing: companies may need less of certain tasks while needing significantly more data engineering, cloud infrastructure, cybersecurity, enterprise architecture, systems integration, governance, program management, and business analysis. They may need specialized expertise during one phase of an initiative, added capacity during another, and long-term internal ownership once the technology reaches production.
The workforce requirement is becoming more dynamic. Many companies are still trying to meet it with a fixed talent model.
Most companies will not build their own foundational AI models. They’ll adopt platforms built by others, connect them to existing applications, provide access to organizational data, and introduce them into established business processes.
That work extends far beyond selecting an AI product or hiring a small number of machine learning specialists.
Data must be prepared, governed, secured, and made accessible. Cloud and infrastructure environments must support new performance, computing, and storage demands. Applications must be connected, tested, monitored, and maintained. Leaders must establish controls around sensitive information, intellectual property, third-party systems, and model behavior. The business must also decide where AI should be used, what decisions require human involvement, how workflows will change, and how employees will be prepared to operate in the new environment.
We see this pattern repeatedly in the teams companies assemble around large-scale initiatives. A priority may be described internally as “an AI project,” but the immediate talent need is a data engineer, cloud architect, cybersecurity specialist, application developer, program manager, or business analyst. These aren’t separate workforce trends running alongside AI. They’re the talent required to implement it. Across the technology staffing market, we consistently see companies pull back in one skill area while demand climbs in cybersecurity, data engineering, and data infrastructure. The public conversation tends to focus on the roles being reduced, without paying equal attention to the work being created around the technology.
The AI platform may get the attention. The surrounding talent determines whether it produces results.
Traditional workforce planning often starts with positions: leaders decide how many developers, engineers, analysts, or project managers they expect to need and then seek approval to fill those roles.
AI requires organizations to begin one step earlier, by identifying the capabilities required across the life of the initiative.
Before an AI application can generate value, the organization may need to modernize data, expand infrastructure, connect systems, strengthen security, redesign workflows, and prepare employees to use the technology. Each phase creates different talent requirements, and those requirements won’t necessarily arrive at the same time or last for the same duration.
Some needs are enduring. A company may decide that enterprise architecture, data governance, cybersecurity leadership, and ownership of its core technology environment should sit inside the permanent workforce; these capabilities are tied closely to accountability, institutional knowledge, and long-term strategy.
Other needs concentrate around implementation. Additional developers may be needed while applications are connected. Data engineers may be needed to build or modernize pipelines. Testing teams may expand as systems move toward production. Project managers and business analysts may be essential during deployment but less necessary once the operating model is established.
Still other roles are simply taking shape. An organization may know it needs a particular capability now, without knowing whether the responsibility will become a permanent position, move into an existing function, or shrink as platforms mature.
A permanent-hiring-only strategy struggles to respond to that variation. An entirely contingent workforce introduces a different risk: leaving the organization without enough ownership or institutional knowledge. The goal isn’t to choose one talent model. It’s to match the model to the work.
Direct hiring remains essential for capabilities that will define the organization’s long-term technology environment.
Companies need internal leaders who understand the business strategy, make architectural decisions, establish governance, and remain accountable after an implementation team has moved on. They also need professionals who understand the company’s systems, data, operating requirements, and industry environment in ways that can’t be quickly transferred to an outside resource.
Direct hire is generally the stronger option when a capability is central to long-term strategy, requires significant institutional knowledge, or involves ongoing ownership of critical systems and decisions. The challenge is that permanent hiring can take time, particularly when organizations are competing for scarce or emerging skills. Leaders should identify these roles early and begin building the internal team while the broader AI strategy is still being developed.
Staff augmentation adds experienced professionals to an existing team for a defined need, period, or stage of work.
This model is especially useful when a company has internal leadership and ownership but lacks sufficient capacity or a specialized skill required to execute the initiative. The technology leader may know where the organization is going and who will ultimately own the environment, but still need additional data engineers, developers, architects, testers, or project managers to reach an important milestone.
Many of the roles surrounding AI lend themselves to this approach because the intensity of the need changes throughout implementation. Data preparation and modernization may require significant capacity early on. Development and integration work may increase in the middle stages. Testing, deployment, training, and adoption support may become more important later. Staff augmentation lets the technology workforce change with the work, rather than forcing the company to build permanent headcount around the busiest phase of the project.
Project-based roles, including program managers, project managers, business analysts, architects, developers, and testers, are often strong candidates for staff augmentation when the organization can define the objective or reasonably anticipate the duration of the work. Highly environment-specific infrastructure positions, by comparison, may be better suited to direct hire when retaining operational knowledge is critical.
The decision shouldn’t be based on job title alone. A developer supporting a defined integration effort creates a different workforce requirement than a developer who will own a mission-critical platform for the next several years.
Contingent staffing gives companies another way to access talent when demand is variable, timing is uncertain, or an emerging capability hasn’t yet been fully defined.
That flexibility is particularly valuable in AI, because the technology and the roles surrounding it continue to evolve. Companies may understand the immediate need without having enough visibility to determine how the work will fit into the future organization.
A contingent professional can help address a current gap while leaders gather more information about the volume, duration, and strategic importance of the work. This is useful when a company is piloting an AI use case, responding to an unexpected increase in project demand, filling a scarce skills gap, or supporting an internal team through a major transition.
Contingent staffing shouldn’t be viewed simply as a faster substitute for permanent hiring. Its strategic value is that it lets leaders align workforce commitments with the amount of visibility they actually have into future demand. When the work is changing quickly, that flexibility has both financial and operational value.
Contract-to-hire bridges the space between an immediate capability need and a potential permanent position.
This is especially relevant for AI-related roles whose responsibilities are still evolving. A company may need someone immediately but remain uncertain about the eventual scope of the position, the level of ongoing demand, or where the role should ultimately sit within the organization.
A contract-to-hire arrangement gives the company time to evaluate the work in a real operating environment, and gives the professional a chance to understand the company, team, technology, and long-term direction of the position. The model shouldn’t be treated as merely an extended interview. It’s a practical workforce-planning tool for situations where the company believes a permanent need may exist, but doesn’t yet have enough information to define it confidently.
The appropriate workforce model will often change as an AI initiative moves from planning through implementation and into long-term operation.
During strategy and planning, the company may need architects, data strategists, business analysts, and security leaders. During implementation, the need may shift toward data engineers, software developers, integration specialists, project managers, and testers. As the technology enters production, the emphasis may move toward permanent system ownership, governance, training, support, and continuous improvement.
A company can use direct hire, staff augmentation, contingent staffing, and contract-to-hire within the same initiative without creating a fragmented workforce strategy. The internal team provides direction, accountability, and long-term ownership. Staff augmentation supplies specialized project capacity. Contingent professionals address fluctuating or emerging needs. Contract-to-hire provides a path for roles that may become permanent. Consulting or project teams take responsibility for clearly defined outcomes requiring coordinated expertise.
This is why, at IDR, we push workforce conversations to start with what the company is trying to accomplish, not with an isolated requisition. The most useful questions aren’t simply whether the organization needs a data engineer or how quickly a developer can start. They’re what stage the initiative is in, which capability is missing, how long the work is likely to continue, what knowledge must remain inside the company, and how much certainty exists about future demand. The answers determine not only whom the company needs, but how that talent should be engaged.
Employers expect 39 percent of workers’ core skills to change by 2030.⁵ The greatest constraint on AI adoption may have less to do with access to technology and more to do with whether organizations can assemble, and continually adapt, the talent required to apply it.
Companies should build the workforce plan alongside the technology, data, infrastructure, security, and operating decisions behind an AI initiative, not after them. That plan should identify which capabilities must be owned permanently, which are needed for a defined period, which can be developed internally, and which require outside expertise. It should also account for how those needs will shift as the initiative progresses.
AI will automate some tasks, create new responsibilities, raise the value of certain skills, and reduce demand for others. What it won’t do is eliminate the need for deliberate workforce planning.
The question isn’t whether your organization needs this talent. It’s whether your hiring model can deliver it fast enough.
Will Hayes is Chief Operating Officer at IDR, where he works with technology leaders to build flexible workforce strategies for AI and digital transformation initiatives. To talk through what a flexible talent model could look like for your organization, reach out to the IDR team at [contact link].