AI is now a cost that changes from one employee to the next, and that makes workforce planning harder to get right. The way through is to plan around the cost of the work, with people and AI together, and HR is where that plan comes together.
For business leaders, the most appealing aspect of AI is its potential to reduce costs. While AI can certainly help improve the bottom line through increased productivity, there is one problem that is impacting AI strategy: the technology is not cheap.
Uber recently discovered this when the company spent its entire 2026 AI budget in a mere four months. Originally, the company had encouraged its engineers to run experiments and maximize their use of the new technology, but the economic reality soon caused a change in strategy. Since then, Uber has revised its policy, with stricter rules about AI access and a monthly spending cap of $1,500 per employee for each AI coding tool.
Not every company has an engineering team as big as Uber’s, but this story does highlight a new business reality. AI usage is an expense, and it’s one that can vary on a per-employee level. When you’re thinking about your workforce budget, you may need to think about how your team uses AI.
The new variable in workforce planning
Traditionally, employers could easily calculate the fully loaded cost of an employee: salary, benefits, payroll taxes, technology, and training. Some people might have flexible costs such as bonuses or travel expenses, but workforce planning was generally quite stable.
AI adds a highly unstable variable to that equation, and it cuts both ways:
-
AI usage: Some organizations are on per-usage billing, which means that heavy AI users can incur large bills. Access to specialist AI tools can also be expensive.
-
AI output: Using AI tools can directly lower costs by allowing people to work faster or streamline processes. In some cases, users might be able to fully automate certain workflows, which can change their workload a great deal.
For example, consider two project managers on the same $100,000 salary. One uses AI occasionally for drafts and meeting summaries. Their AI costs are negligible, but so is the extra output they get from it.
The other project manager might use agents for planning, communicating, and automatically responding to incidents. A heavy AI user might incur substantial additional costs, but this might be a solid investment if it means they’re producing better results than their colleague.
That means business leaders need to shift from thinking about the cost of an employee to the cost of the work. This figure is the combined cost of everything required to get the work done, both people and AI.
What this means for HR and workforce planning
This is where AI becomes an HR strategy issue. Workforce planning is a core part of business strategy, and that planning becomes much more difficult when you can’t accurately gauge the cost of work. That’s why it’s important to work through all of the factors that could impact planning.
1. Look for AI champions
AI usage varies greatly between individuals. Some people might only use AI tools for simple tasks like drafting emails, while others might have already automated large parts of their day-to-day work. These champions are an invaluable source of information, as they can tell you exactly where AI helps and where it creates new challenges.
These insights will help give a sense of how AI might make a real difference to your operations. You can also invite your team to offer new ideas for ways that they could use automation in their roles. How could AI help them cut out repetitive work, or offer improved service to clients?
2. Target high-impact gains
AI can help your team do many things, but which of those things will change your workforce planning needs? Before you increase AI usage, test each gain against three factors:
-
Impact: AI usage should produce a result you can measure, whether that’s increased productivity, improved customer experience, or the ability to provide new types of services.
-
Reliability: A poorly designed AI process can create errors and rework, which adds to overall workforce costs.
-
Scalability: If an AI workflow works for one team, you can multiply the gain by rolling it out to others. That only works if the workflow is something other people can learn and follow, and doesn’t depend on one enthusiastic employee.
The gains that meet all three tests are the ones worth building your workforce plan around.
3. Review role requirements
Once you know where AI adds value, go back to the roles themselves. Work through each job description with the manager and sort every responsibility into one of three groups:
-
Automate: Tasks AI can handle from start to finish with a light human check, such as scheduling, first-draft reports or routine data entry.
-
Augment: Tasks where AI makes the person faster or better but the judgment stays with them, such as analysis, planning or client communication.
-
Unaffected: Tasks that depend on relationships, physical presence or judgment that AI can’t yet support.
This is a chance to restructure some jobs. If much of a role falls under “automate,” the freed-up time could go to higher-value work, or the role could merge with another. Either way, update the job description to match. Only 16% of mid-market employers have updated job descriptions for AI, and outdated ones cause problems in hiring, reviews, pay decisions and FLSA exempt classification.
4. Connect the AI budget to the workforce budget
If you’re building your strategy around a cost-of-work model, you’ll need to connect two budgets. AI spend usually sits in the IT budget, while salaries sit in the workforce budget. This is the point where you’ll really see how AI-supported employees will affect the bottom line.
It’s crucial to consider how AI pricing might affect your overall plans. Some providers charge per seat, which offers a fixed cost but may include usage limits. Others bill on a per-usage basis, which means greater access but less predictable costs. Some employees may also require access to specialist AI tools or services, which can also increase costs.
5. Upskill or hire to build the right team
The best way to get a return on your AI investment is to have a team that understands the tools. The first step here is to implement an upskilling program that helps bring your employees up to speed. This is also a good chance to educate them on appropriate usage, including cost management on expensive AI tools.
Upskilling won’t close every gap. Some work depends on judgment, relationships or specialist skills your team doesn’t have yet, and that’s where hiring still makes sense. When you do hire, recruit against the updated job description so the new person starts with the way the work is done now, including the AI skills the role needs.
Start with the work
You don’t need to predict what every job will look like in five years. You do need a clear view of how work gets done today and where AI could change it. With that in hand, decisions about skills, staffing and AI spending can support each other, and you’ll know where spending more on AI pays off and where it doesn’t.
Helios HR can help you plan for the people side of AI:
-
Strategic HR for organizations that need a workforce plan covering people, skills and AI capacity together
-
Performance management for organizations resetting goals and expectations as AI changes the work
-
Training and development for organizations whose people have AI tools but not yet the skills to use them well
-
HR compliance for organizations redesigning roles and updating job descriptions around AI
Contact Helios HR to plan for AI in your workforce today.
Frequently asked questions
Is AI actually cheaper than employees?
Sometimes, for specific tasks. AI can handle routine work like scheduling, first drafts and data entry at low cost. But heavy usage, specialist tools and agents add up quickly, and most work still needs human judgment. The useful comparison is the total cost of getting the work done.
How much does AI cost a business?
It depends on how you pay and how your people use it. Per-seat plans give a fixed monthly cost, often with usage limits. Usage-based billing gives more access but less predictable bills, and costs can vary widely from one employee to the next.
What is the cost of work?
It's everything it takes to get a piece of work done. That combines people costs such as salary, benefits and training with the AI tools and usage that support them. It lets you compare hiring, upskilling and AI spending on equal terms.
How does AI change workforce planning?
AI adds a new variable to the plan. When a team needs more capacity, the answer might be a new hire, more AI support for existing staff, redesigned roles or a mix. Good planning weighs those options together against the outcomes you need.
Should companies cap employee AI spending?
Caps can stop runaway bills, but they work best as a backstop. Some roles get far more value from AI than others, so it's a good idea to match access to the work each role does and to measure what the spending produces.
Is it worth paying for AI?
It's worth it where AI produces a result you can measure, such as faster delivery, fewer errors or better client service, and where the workflow is reliable and repeatable. Test each gain before you scale it, so you know which spending pays off.
What role should HR play in AI decisions?
HR brings the workforce view. IT understands the tools and Finance the costs, while HR can answer what AI means for roles, skills, performance expectations and staffing. That makes HR a natural partner in deciding where AI spending pays off.