AI Literacy at Work: What Your People Need to Know and What the Law Now Expects

Since 2 February 2025, the EU AI Act has required every organization that builds or uses AI systems to take measures to support the AI literacy of its staff. The rule was reworded in July 2026, national authorities began supervising it in August 2026, and in the United States the Department of Labor published its own AI Literacy Framework in February 2026. AI literacy has moved from a training-catalog topic to something regulators name directly. This guide explains what the term means, what the EU and US texts actually ask for, which skills matter by role, and why a training course on its own rarely produces it. It is general information, not legal advice.

What does AI literacy mean?

AI literacy is the practical ability to use AI tools well, judge what they produce, and understand their risks. It is not the ability to build models or write code. The EU AI Act defines it as the “skills, knowledge and understanding” that allow people “to make an informed deployment of AI systems, as well as to gain awareness about the opportunities and risks of AI and possible harm it can cause.”

The US Department of Labor’s framework describes it in similar terms, as a foundational set of competencies that enable individuals to use and evaluate AI technologies responsibly. Both definitions have the same three parts: using AI, evaluating it, and knowing where it can go wrong. An employee who can get a useful draft out of an AI assistant but cannot tell when the draft is wrong has only the first third.

Is AI literacy a legal requirement?

In the European Union, yes. Article 4 of the AI Act says providers and deployers of AI systems “shall take measures to support the development of AI literacy of their staff and other persons dealing with the operation and use of AI systems on their behalf.” In the United States there is no equivalent federal obligation; the Department of Labor’s framework is voluntary guidance.

Three details of the EU rule matter in practice:

• It covers users of AI, not only makers. A “deployer” is any person or organization using an AI system under its authority, other than for personal, non-professional use. A company whose staff use an AI assistant at work is a deployer even if it has never built a model.
• It reaches beyond employees. The European Commission’s guidance says “other persons” includes people broadly under the organization’s remit, such as a contractor, a service provider, or a client.
• It was softened in 2026, not removed. The amended text adds that the obligation “does not require providers or deployers to guarantee any specific level of AI literacy of any individual.” The duty to take measures remains.

	EU AI Act, Article 4	US Department of Labor AI Literacy Framework
Status	Legal obligation, in application since 2 February 2025, amended July 2026	Voluntary guidance, issued 13 February 2026
Who it addresses	Providers and deployers of AI systems	The public workforce and education systems, as a resource for program design
What it asks	Take measures to support AI literacy, taking into account people’s knowledge, experience, and the context of use	Encourages expanded AI literacy training built around five content areas
Proof expected	No certificate needed; the Commission says internal records of training and other initiatives are enough	None
Enforcement	National market surveillance authorities, from August 2026; penalties set nationally and must be proportionate	None

The Commission is explicit that there is no prescribed course: “no strict requirements or mandatory trainings are imposed,” and “there is no one size fit all when it comes to AI literacy.” What an organization does should depend on its role, the risk of the systems it uses, and what its people already know.

What should an AI-literate employee be able to do?

An AI-literate employee can explain in plain terms how the tools they use work, pick the right tasks for them, give clear instructions, check the output, and use them within the rules. The US Department of Labor’s framework organizes this into five content areas, which make a practical checklist for any workplace.

1. Understanding AI principles. Knowing that a language model predicts likely text and does not look facts up, which is why it can be fluent and wrong at the same time.
2. Exploring AI uses. Recognizing which parts of a job AI helps with, such as first drafts, summaries, and data preparation, and which it does not.
3. Directing AI effectively. Giving context, constraints, and examples, and breaking a task into steps the tool can handle.
4. Evaluating AI outputs. Checking facts, figures, and sources before anything is used, and knowing when a result needs an expert’s review.
5. Using AI responsibly. Knowing what data must never be entered into a tool, when to disclose AI use, and who is accountable for the result.

The fourth area is where most organizations are weakest and where most of the risk sits. Teaching people to prompt is quick. Teaching them to doubt a confident answer takes practice on real work.

Does everyone need the same level of AI literacy?

No. The EU text itself says measures should take into account people’s “technical knowledge, experience, education and training and the context the AI systems are to be used in.” A sensible program sets a common baseline for everyone and adds depth by role.

Group	What they need
All staff	What AI can and cannot do, the company’s usage rules, what data is off limits, and how to check an answer
Daily users	Task-specific practice in their own workflows, verification habits, and when to escalate
Managers	How to review AI-assisted work, set expectations for their team, and answer day-to-day questions
Owners of AI systems and workflows	Risk assessment, testing, monitoring, and the obligations that apply to the systems they run
Senior leaders	Enough understanding to set priorities, fund the right use cases, and be accountable for governance

Managers deserve more attention than they usually get. McKinsey’s 2025 workplace research found that two-thirds of managers field questions from their team about how to use AI tools at least once a week. In most companies they are already the AI trainers, whether or not anyone has trained them.

Why is training alone not enough?

Training gives people knowledge. AI literacy shows up only when that knowledge is used in daily work, and that depends on clear rules, redesigned tasks, and incentives as much as on courses. The demand for training is real, but the gap between what employees do and what leaders think they do is wider than a course can close.

McKinsey found that C-suite leaders estimate only 4% of employees use generative AI for at least 30% of their daily work, when employees report a share three times greater. Nearly half of employees say they want more formal training and believe it is the best way to boost AI adoption, yet more than a fifth report minimal to no support. People are using AI faster than their organizations are teaching them to use it safely.

Infomineo’s AI Scaling Framework treats this as one of seven building blocks for scaling AI, called workforce activation: equipping employees with the skills, incentives, and clarity needed to integrate AI into daily work. It breaks the block into four capabilities, and only the first is training:

• Role-based training: tailored AI training aligned with specific roles and responsibilities.
• Usage clarity: defining when and how employees should use AI tools in their daily workflows.
• AI-native job redesign: restructuring roles to integrate AI as a core component of task execution.
• Incentive alignment: aligning performance metrics and rewards with effective AI adoption and outcomes.

An organization that runs a course and stops there has done one of the four. The other three are why AI adoption stalls even in companies that have trained everyone, and why workforce readiness is a core part of AI scalability.

How do you build an AI literacy program?

Start from the AI your organization actually uses and the people who use it, set a baseline by role, teach on real work, and keep a record. The European Commission suggests four considerations as a minimum: a general understanding of AI in the organization, the organization’s role as provider or deployer, the risk of the systems involved, and tailoring to staff knowledge and the context of use.

1. List the AI systems in use and who uses them. Include tools staff adopted on their own. You cannot scope literacy for systems you have not identified.
2. Sort them by risk. A drafting assistant and a system that screens job applicants do not call for the same depth of understanding.
3. Set the baseline and the role-specific layers. Use the role table above as a starting point and adjust it to your organization.
4. Write down the usage rules. What tools are approved, what data may be entered, what must be checked, and who signs off. Rules people can find are part of literacy.
5. Teach in real workflows. Practice on the tasks people do every day sticks better than a general course. The Department of Labor’s framework makes the same point with its principles of experiential learning and embedding learning in context.
6. Equip managers first. They are the ones answering questions every week.
7. Keep records and refresh. Note who was trained on what and when. Tools and rules change, so revisit the program at least yearly. In the EU, these internal records are what the Commission says is sufficient evidence.

If you are unsure where your organization stands, an AI readiness assessment is a useful first step, since it covers skills alongside data, technology, and governance.

Infomineo’s AI and Data Advisory gives you end-to-end guidance in developing and implementing AI and Data strategies, identifying high-impact use cases where Human-AI synergy offers the greatest leverage.

Talk to our AI and Data Advisory team →

Frequently Asked Questions

What is AI literacy in simple terms?

AI literacy is knowing how to use AI tools well, how to check what they produce, and what can go wrong. It does not require coding or building models. The EU AI Act defines it as the skills, knowledge, and understanding needed to make an informed deployment of AI systems and to be aware of their opportunities and risks.

Does the EU AI Act require AI literacy training?

It requires providers and deployers of AI systems to take measures to support the AI literacy of staff and others using AI on their behalf. It does not prescribe a specific course or a certificate. The European Commission says there are no mandatory trainings and that internal records of what was done are sufficient.

Does the EU AI literacy rule apply to companies outside the EU?

It can. The AI Act applies to providers placing AI systems on the EU market wherever they are established, and to providers and deployers located outside the EU where the output produced by the AI system is used in the Union. Whether Article 4 applies to a specific company depends on its role and activities, so confirm with legal counsel.

Is AI literacy the same as AI training?

No. Training is one way to build AI literacy. Literacy is the result: people who use AI well and safely in their daily work. That also depends on clear usage rules, tasks redesigned around the tools, and incentives that reward good use.

Who needs AI literacy in a company?

Everyone who uses AI or acts on its output, at different depths. All staff need a baseline. Daily users, managers, owners of AI systems, and senior leaders each need more, matched to the decisions they make and the risk of the systems they work with.