EU AI Act: Myths and Facts

The most common myth about the EU AI Act is that it only applies to big tech giants and AI developers - this is untrue, as the key AI literacy obligation (Article 4) applies to all companies using AI, regardless of size, and has been in effect since 2 February 2025. In Lithuania, many misconceptions have accumulated around Regulation (EU) 2024/1689: some believe nothing needs to be done, while others fear expensive lawyers and months of work. In practice, what is usually needed is less panic and more systematic evidence.

Below, we have compiled the most common AI Act myths alongside facts, supported by the regulation itself. The goal is to clearly separate what truly applies to you from what is merely online fear or sales gimmicks. If you are not yet familiar with the topic, start by reading EU AI Act Explained Simply, then return here.

Myth: "The EU AI Act only applies to large companies"

Fact: The AI literacy obligation applies to all companies using AI - from one person to a thousand.

This is perhaps the most harmful misconception. Article 4 of the EU AI Act, amended on 27 July 2026, requires that AI system providers and deployers (i.e., users) take measures to support the development of AI literacy among their staff and relevant persons, considering their knowledge, experience, training, context, and the people affected. There is no exemption for "companies with fewer than X employees". A micro-enterprise where three people use "ChatGPT" for text generation or AI translation falls under the same Article 4 obligation as a corporation.

What is true is the principle of proportionality. Some more complex obligations (e.g., high-risk system requirements) are tailored to the size of the company and the risk involved. However, the AI literacy obligation itself does not differentiate by size. For more details, see Who Does the EU AI Act Apply To?.

Myth: "We don’t use AI, so it doesn’t apply to us"

Fact: Most companies use AI without even realising it.

When you ask the manager of a small Lithuanian company if they use AI, the answer is often "no". But in reality, AI is already embedded in everyday tools:

• Text writing and editing with "ChatGPT", "Gemini", or "Copilot";
• AI translation (e.g., "DeepL");
• Customer chatbots on websites;
• Accounting and invoicing automation with AI features;
• Image and advertisement generation;
• Email writing "assistants" and browser extensions.

If even one employee uses any of these tools for work purposes, your company is an AI deployer. This is not a bad or prohibited thing; it simply means that the AI literacy obligation applies. The first practical step is to openly list all the tools used, including the "invisible" ones.

Myth: "Everything only takes effect in 2026"

Fact: The most important obligation for small businesses has been in effect since 2 February 2025, with supervision and enforcement starting in August 2026.

Many delay action, believing they have time until August 2026. This is a dangerous mistake, as the AI literacy obligation and prohibited AI practices have been applicable since 2 February 2025. Here is an accurate overview of the dates:

In short, the "2026" myth obscures what is already mandatory. For most micro-enterprises, the key date has passed - 2 February 2025. This does not mean panic, but it does mean there is no time to delay.

Myth: "There are no penalties, this is just a formality"

Fact: The AI Act provides for real, significant fines - but for small businesses, the risk is often practical rather than just fines.

The regulation sets out administrative fines depending on the severity of the infringement: for prohibited AI practices, up to €35 million or 7% of global annual turnover (whichever is higher); for other violations, lower but still substantial amounts. For micro-enterprises, the principle of proportionality applies, so fines are adjusted to the company’s capacity.

However, for small businesses, the fine itself is often less important than the indirect risks: loss of customer or partner trust, issues in public procurement or with larger clients’ compliance checks, or personal data leaks due to careless AI use. We discuss these risks in more detail in the article on AI Act Fines and Risks. The myth "there are no penalties" is dangerous because it ignores this entire layer of practical risk.

Myth: "You’ll need expensive lawyers and months of work"

Fact: Proportionate guidance and records can help micro-enterprises, but what is needed depends on their use of AI.

This myth is often spread by those looking to sell expensive consultations. The reality is simpler: a typical Lithuanian micro-enterprise uses low or limited-risk AI tools (text generation, translation, images), so it does not need complex compliance assessment procedures designed for high-risk systems.

What is actually needed for most:

• AI Usage Policy - internal rules on what is and isn’t allowed with AI;
• AI Tools Register - a list of AI tools the company uses and their associated risks;
• Compliance Record - evidence that Article 4 obligations are being fulfilled;
• Employee AI Guide - practical guidelines for the team;
• AI Literacy Training with knowledge checks and certificates.

When all of this can be prepared using standard templates, and when a lawyer is needed, we explain in the article Do You Need a Lawyer for AI Compliance?. For example, nebedirbam.lt automatically prepares this entire Article 4 documentation package based on expert-reviewed templates for a fixed €149 + VAT (€299), without consultation queues or hourly fees. These documents and training are a recommended practical approach, not a mandatory legal minimum or a guarantee of compliance. If you first want to try the training and obtain your own completion certificate, you can start with free AI literacy training. The free path is for one signed-in participant; team training is included in the paid documentation package.

Myth: "AI literacy is the same as documentation"

Fact: Learning and documentation serve different purposes. Article 4 requires measures supporting literacy, not a prescribed document set.

Article 4 of the EU AI Act requires AI literacy - meaning people’s knowledge and skills to use AI responsibly. Documentation (policy, register, compliance record, guide) is evidence that you are fulfilling these obligations and controlling AI use.

Training and guidance support learning, while records help explain the steps taken. Possessing documents does not replace actual measures. Clear records of learning activities can be useful when reviewing practice or answering questions. Article 4 does not prescribe a specific document bundle.

Myth: "A one-time fix is enough, then you can forget about it"

Fact: Compliance is an ongoing process because tools change.

Companies start using new AI tools, update old ones, and employees change. If a year ago you listed your AI tools, but since then the team has started using three new ones, the documentation no longer reflects reality. Compliance is not a one-time "checkbox", but a continuous habit of updating.

In practice, this does not mean a lot of work: it is enough to review the documents when a new significant AI tool is introduced or when regulatory obligations change. The key is to have a clear structure that is easy to update, rather than rewriting everything from scratch each time.

Quick Summary: Myths vs. Facts

Where to Verify the Facts?

The best way to avoid falling for myths is to check the primary source. The official text of the regulation in all EU languages can be found on the EUR-Lex portal (eur-lex.europa.eu) by searching for "Regulation (EU) 2024/1689". Additional explanatory guidance and FAQs are provided by the European Commission’s AI Act Service Desk page ai-act-service-desk.ec.europa.eu and the Digital Strategy page digital-strategy.ec.europa.eu. If someone online claims something about the AI Act and you are unsure, verify it there.

Want to dive deeper? Start with EU AI Act Explained Simply, review Who the Regulation Applies To, and assess AI Act Fines and Risks. All articles on this topic can be found in the article list.

Conclusion

Many misconceptions have accumulated around the EU AI Act, which harm small businesses in two ways: some myths encourage inaction, while others unnecessarily scare with expensive processes. The truth is simple: the AI literacy obligation (Article 4) applies to all companies using AI and has been in effect since 2 February 2025. Fines are real, but for most micro-enterprises, proportionate, standardised documentation and employee training are sufficient. The requirement is to take measures to support AI literacy, but there is no universal mandate for specific certificates, a fixed set of documents, training hours, external trainers, audits, or a dedicated AI officer. Our documents and quizzes are recommended practical tools, not a legal minimum or compliance guarantee. Separate myths from facts, verify with primary sources, and address compliance without panic: it can be done quickly with standardised tools.

The facts in this article are verified against official European Commission sources: AI Act Service Desk (Article 4) and Digital Strategy Page.