67% of HR Leaders Don't Know What AI Can Actually Do. SHRM's 2026 Data Says the Real CHRO Bottleneck Isn't Budget

The budget excuse is gone. The compliance excuse is gone. According to the Society for Human Resource Management's (SHRM's) 2026 State of AI in HR report, the number one reason HR has not adopted AI is that HR leaders don't know what it can do.

That's a harder problem to solve than a budget line item.

What SHRM's 2026 Data Actually Says

SHRM surveyed 1,908 HR professionals in December 2025. The findings landed with unusual clarity for a research report that covers a topic as noisy as artificial intelligence and human resources.

Key Facts

• 92% of CHROs expect AI to be further integrated into the workforce this year (SHRM, 2026).
• 67% of HR leaders say "we don't know what AI can do" is the number one reason they haven't adopted AI in HR (SHRM, 2026).
• 62% of organizations deploy AI somewhere, but only 39% deploy it inside HR (SHRM, 2026).

Ninety-two percent of chief human resources officers (CHROs) expect AI will be further integrated into the workforce this year. Eighty-seven percent forecast greater AI adoption inside HR processes specifically, up from 83% in 2025. The expectation side of the equation is near-universal.

The adoption side tells a different story. Sixty-two percent of organizations are deploying AI somewhere in their operations. But only 39% have deployed it inside the HR function. That gap between organizational AI deployment and HR-specific AI deployment points to something structural, not situational.

And the SHRM data names it directly. When HR leaders were asked why they haven't adopted AI in HR, 67% cited a single reason: we don't know what it can actually do. Not budget. Not data privacy. Not lack of executive support. Awareness is the primary blocker.

For reference, recruiting is the leading AI use case in HR right now, active at just 27% of organizations. So the function that processes, evaluates, and develops human talent at scale is largely operating without the technology it's been told for two years will reshape everything -- and the reason is that the people responsible for that function can't yet describe what they'd be adopting.

Why This Matters More Than the Mercer Readiness Gap

If you've been tracking the 2026 CHRO data cycle, you've seen Mercer's numbers showing 98% of executives acknowledging AI's importance while only 50% feel ready to lead through it. LinkedIn's fastest-growing roles report shows AI-adjacent positions climbing faster than any other category. The ICIMS hiring data flags a collapse in entry-level opportunities as AI absorbs early-career tasks.

All of those data points share a quiet assumption: that HR leaders understand what AI is and can do in an HR context. The Mercer readiness gap assumes you know what you're not ready for. LinkedIn's skills data assumes HR can evaluate whether a candidate's AI skills are genuine. The ICIMS entry-level story assumes HR is in a position to redesign the pipeline with intention.

SHRM's finding breaks that assumption for two-thirds of the profession.

Think about what it means to run a request for proposal (RFP) process for an HR technology platform when 67% of the evaluation team can't distinguish a genuine AI feature from a marketing label. Vendors know this. They use terms like "AI-powered," "intelligent matching," and "predictive analytics" in product collateral without specifying what model is running, what data it was trained on, how it handles bias, or what it actually predicts. If the buyers can't push back on vague claims, there's no incentive to provide precision.

This is the practical consequence of the awareness gap: procurement processes select for marketing confidence rather than technical substance. And the organizations that win those RFPs aren't necessarily deploying better AI -- they're deploying better positioning.

The AI workforce readiness gap explored in the 2026 CHRO context is related but distinct. Workforce readiness asks whether employees can work alongside AI. The SHRM data asks something prior: can HR leadership evaluate the AI that employees will supposedly be working alongside? You can't design a readiness program around a system you can't name.

The HR-AI Literacy Ladder

The most useful reframe for CHROs right now isn't "how do we adopt more AI" -- it's "where does our function sit on the literacy curve, and what does one rung up look like?"

Here's a framework that maps the terrain:

The HR-AI Literacy Ladder describes four levels of functional AI understanding for HR leaders:

Level 1: Aware. The HR leader knows AI exists in HR tools. They've seen demos. They've read the analyst reports. They can confirm, in a board meeting, that AI is reshaping talent acquisition. But they couldn't tell you which specific tasks AI is handling in any given product, or how to verify a vendor's claim.

Level 2: Discerning. The HR leader can tell a real AI feature from a marketing label. They understand the difference between a rules-based chatbot and a large language model. They know to ask what the model was trained on and what "AI-powered screening" actually filters for.

Level 3: Specifying. The HR leader can write a useful AI requirement into an RFP. They can articulate what outcome they want, what data is available to train or tune the model, what bias audit they expect, and what success looks like in the first 90 days.

Level 4: Operating. The HR leader can run a pilot and judge whether it worked. They can compare pre- and post-deployment metrics, identify confounding variables, and decide whether to expand, adjust, or stop.

SHRM's data puts most HR leaders between Level 1 and Level 2. They're aware AI is coming -- 92% expectation levels confirm that. But they can't yet tell whether a specific tool's AI feature is substantive or cosmetic. That single rung gap, from Aware to Discerning, is worth more to a CHRO in 2026 than any single tool purchase.

You can buy a great product and still get nothing from it if you can't specify what you need or evaluate what you got. The HR-AI Literacy Ladder isn't just a diagnostic -- it's a purchasing filter.

For deeper context on how AI literacy connects to strategic leadership capacity, see leading AI transformation at the organizational level and why most AI transformations fail.

The Governance Problem Hidden in the Data

There's a second implication that's easy to miss. CHROs are increasingly being asked to lead AI governance -- to own ethics frameworks, bias audits, and workforce impact assessments. That work is being staffed with professionals who, per SHRM's data, largely can't yet articulate what the systems they're governing actually do.

That's not a criticism of CHROs as individuals. It's a description of where the profession is in a very fast-moving cycle. But it has real consequences. An AI governance framework built without functional AI literacy is a framework built around assumptions, not mechanics. It governs the idea of AI rather than the practice of it.

The LinkedIn 2026 data on self-upskilling gaps for AI engineers and HR professionals suggests that even in functions with more AI proximity than HR, the support infrastructure for closing skill gaps is thin. For HR, which is still largely at Level 1 on the literacy ladder, the gap is wider and the support is thinner.

This connects directly to the question of AI experimentation versus execution maturity: organizations that experiment without a discerning buyer class in HR tend to buy broadly and deploy narrowly, which produces the exact 62%/39% gap SHRM is documenting.

What to Do This Quarter

The SHRM data is a diagnosis, not a sentence. Closing one rung on the literacy ladder is achievable in a quarter if the work is structured. Here's what that looks like:

1. Run a 3-vendor AI discovery sprint. Bring in three HR tech vendors -- one you already use, one you're evaluating, one you've never heard of. Ask each to walk your team through exactly one AI feature: what it does, what data it uses, how it handles edge cases, and how they'd measure its performance. Don't evaluate the product. Evaluate your team's ability to ask the right questions. Document the gaps.

2. Rewrite one active RFP with an AI specification. Take a current or upcoming HR technology RFP and add a dedicated AI requirements section. Force specificity: what task does the AI perform, what data trains it, what's the bias audit process, what does success look like at 90 days? The act of writing it is the literacy exercise. You'll learn more from drafting the spec than from any vendor demo.

3. Run an internal AI show-and-tell. Ask two or three managers or human resource business partners (HRBPs) who are already using AI tools -- in any function -- to demonstrate what they're doing for 20 minutes each. Make this a monthly standing meeting, not a one-off. Awareness compounds. The people who know something teach the people who don't.

4. Build a basic literacy dashboard. Track where your HRBPs sit on the literacy ladder. A simple self-assessment -- can you name three AI use cases in your function, can you describe one AI feature you use in your current tools, have you run or participated in an AI pilot -- gives you a baseline. Once you have a baseline, you can move it. For a framework on how skills-based approaches apply to internal HR capability building, the principle transfers directly.

The leadership readiness gap in AI is real across functions, but HR's version is particularly urgent because HR is simultaneously the governance owner and the late adopter. Closing the awareness gap inside the HR function isn't optional preparation -- it's the prerequisite for everything else.

FAQ

What does "AI literacy" mean for an HR leader in 2026?

AI literacy for an HR leader isn't about coding or data science. It means being able to evaluate vendor claims (can you tell a real AI feature from a marketing label?), write a meaningful AI requirement into an RFP, and judge whether a pilot produced a result worth scaling. SHRM's 2026 data suggests most HR leaders are at Level 1 of this spectrum: aware that AI is coming, but not yet able to distinguish substance from positioning.

Why isn't budget the number one barrier to AI adoption in HR?

SHRM's 2026 survey of 1,908 HR professionals found that 67% named "we don't know what AI can actually do" as their primary reason for not adopting AI in HR. Budget, compliance, and executive support ranked lower. This is notable because budget constraints are solvable with a business case. Awareness gaps require a different kind of investment: deliberate exposure, structured evaluation, and internal literacy-building that takes time even when money is available.

How can a CHRO close the AI awareness gap in one quarter?

Four moves: run a 3-vendor discovery sprint focused on learning to ask better questions (not on selecting a product); rewrite one active RFP to include specific AI requirements; set up a monthly internal show-and-tell where HRBPs who use AI tools demonstrate them to peers; and build a simple literacy baseline for your HRBP population. These don't require budget approval or an AI strategy document. They require a calendar commitment. The 5 stages of AI maturity framework maps what organizations look like at each stage -- knowing where HR sits helps CHROs sequence the work.

Learn More

• What is business AI: a primer for non-technical leaders
• AI governance framework for executives
• Leading AI transformation at the organizational level
• The 5 stages of AI maturity
• AI workforce readiness gap 2026 (CHRO edition)