What the DOL's AI Literacy Framework Means for Employers

Most employers running AI training programs today are doing so without a clear external benchmark to measure against. They are making reasonable decisions in the absence of a shared standard: which tools to cover, which employees to prioritise, what good looks like.

The numbers confirm the problem. 82% of enterprise leaders say their organisation provides some form of AI training in 2026 — and yet 59% still report an AI skills gap. Training is happening. Capability is not following. 42% of employees say their employer expects them to learn AI on their own, while only 17% use AI tools frequently in their work. A gap that reflects the absence of any shared standard for what AI-ready looks like.

The US Department of Labor's AI Literacy Framework, published on February 13, 2026 through Training and Employment Notice 07-25, changes that. It is not a mandate. No enforcement mechanism exists behind it. But it establishes, for the first time at a federal level, a clear definition of what AI literacy means for workers. It signals the direction of travel for employers who want to stay ahead of workforce expectations rather than catch up to them. This article explains what the framework actually says, what it asks of employers, and what the practical difference is between organisations that treat it as background reading and those that use it.

What the Framework Is — and What It Is Not

The DOL's AI Literacy Framework is voluntary guidance, not regulation. It does not impose obligations on private employers, carry penalties for non-compliance, or require any reporting. Organisations can ignore it without legal consequence.

That framing matters, because the temptation when a government body releases a voluntary framework is to file it and move on. That would be a mistake. Not for compliance reasons. For competitive ones.

What the framework actually is: a federal-level definition of AI literacy developed in response to a straightforward problem. AI is reshaping how work gets done across virtually every sector, and the country lacked a shared vocabulary or common structure for training workers to use it. Before this release, AI literacy efforts across the US operated in largely fragmented fashion: community colleges developed their own curricula, private bootcamps marketed proprietary programs, and state workforce agencies ran pilots that rarely scaled. The framework provides a common architecture for the first time.

Employers who align their training programs to it are not meeting a legal obligation. They are building against the benchmark that hiring, procurement, and workforce development conversations will increasingly reference. Secretary of Labor Lori Chavez-DeRemer said at launch: "The Department of Labor is committed to making sure all American workers are able to share in the prosperity that AI will create for our economy."

What the Framework Defines as AI Literacy

The DOL defines AI literacy as a foundational set of competencies that enable individuals to use and evaluate AI technologies responsibly, with a primary focus on generative AI as the technology most central to the modern workplace.

That definition has a deliberate scope. The framework is not describing what AI engineers or data scientists need. It is describing what every worker needs: the baseline capability to engage with AI tools confidently, appropriately, and without creating risk for the organisation or themselves. The framework organises this into five foundational content areas:

What is notable about that content list is what it excludes. Technical proficiency — building models, writing code, configuring systems — is explicitly out of scope. The most in-demand AI skills in 2026 are not deeply technical. They are interpretive, applied, and judgment-driven. The framework is asking for judgment, not expertise. That distinction matters enormously for how employers design and target their training investment, and it is the same distinction that separates awareness training from genuine upskilling.

What It Says Specifically About Employers

The framework addresses employers directly, and the guidance it offers is more operational than most voluntary frameworks tend to be. Rather than abstract principles, it describes specific decisions employers should make.

It asks employers to start by reviewing current workflows where AI tools are already emerging: drafting, data analysis, customer communication. The framework asks them to assess how basic AI literacy helps employees work more effectively in those specific contexts. Critically, it explicitly asks employers to identify what level of AI literacy different roles require, rather than applying a single standard across the workforce.

This role-differentiated approach is one of the framework's most important practical signals. Nearly a quarter of enterprise leaders cite untailored learning paths as a key failure of their current training programs. The DOL is not suggesting that a warehouse operative and a senior analyst need identical AI literacy training. It is asking employers to do the work of mapping what each role actually requires, and then building training that matches that map. Understanding AI training requirements by role is the starting point for that mapping exercise.

The framework also asks employers to provide clear internal guidance on appropriate AI use, to encourage hands-on practice built around real workplace tasks, and to identify roles that may require deeper proficiency beyond baseline literacy. Bayer's Data Academy illustrates what this looks like at scale: the company built a multi-tier program to strengthen foundational digital and AI fluency across the enterprise, with over 90% of learners reporting improved innovation or processes after completing training. Each of the framework's asks maps directly to program design decisions that most organisations have not yet made explicitly.

The Seven Delivery Principles — and Why They Challenge Most Current Programs

The most practically useful part of the framework for employers is not the content areas. It is the seven delivery principles, which describe how AI literacy training should be structured to actually produce capability rather than completion rates. Read against the most common AI training format in 2026 — a combination of online-based modules and occasional instructor-led sessions — they constitute a direct challenge to what most organisations are currently doing.

The pattern is consistent across enterprise research: passive training formats consistently struggle to build applied AI capability. The framework is not describing what most organisations are doing. It is describing what they should be doing instead. The gap between the two is where most AI skills gaps live.

How This Compares to the EU AI Act's Training Obligations

Employers operating across both US and EU jurisdictions are now navigating two distinct frameworks, with a critical structural difference between them.

The practical implication for multinational employers is that the EU obligation sets the legal floor, while the DOL framework provides a useful content structure for meeting it across a US workforce. The five content areas the DOL defines are substantively consistent with what Article 4 requires in terms of contextual, role-appropriate literacy.

Organisations that build training programs aligned to the DOL framework are not inadvertently meeting the EU AI Act obligation — but they are much closer to it than organisations running generic awareness sessions. For companies with employees in both jurisdictions, the DOL framework is a practical starting architecture for a single program that goes a long way toward satisfying both. The full picture of EU deployer obligations is in the EU AI Act deployer obligations guide, and the Article 4 requirements in detail in what the EU AI Act means for your team's training.

What "Voluntary" Actually Means for Employer Risk

The voluntary nature of the framework does not mean it is consequence-free to ignore. Three practical risks apply to employers who treat it as irrelevant.

Workforce expectations. 55% of employees say access to AI training or certification would make them more likely to stay with an employer, and when employers provide AI training, AI adoption jumps to 76% compared to just 25% without support. As the DOL framework filters into hiring conversations and workforce development programs, employees will increasingly know what AI literacy means, and will notice when their employer's training falls short of it.

Procurement and contracting. Organisations bidding for federal contracts or partnering with public institutions will find AI literacy expectations embedded in those conversations at an increasing rate. The framework provides a shared vocabulary that these conversations will use.

Regulatory trajectory. Voluntary frameworks consistently precede mandatory ones. The current administration has already proposed removing state-level AI laws in favour of a national standard, a direction that makes the DOL framework's definitions more likely to be codified over time, not less. Treating it as an early signal of where US policy is heading is more accurate than treating it as a permanent opt-out.

Three Things to Do With the Framework Now

Voluntary guidance is most useful when it produces specific decisions, not general awareness. Here are the three most actionable things employers can take from the DOL framework.