Plain language at the touch of a button: can AI do it?

Why understandable texts matter

Many people find it hard to understand complex texts: people with learning difficulties or cognitive impairments, people with limited reading skills, people who are learning the language and many older people. According to the latest OECD Survey of Adult Skills, published in December 2024, 26% of adults aged 16 to 65 in the participating countries – most of them EU Member States – read at the lowest level or below.

For all of them, an official letter, a product description or a form full of jargon and legal language can become a barrier. Understandable texts are therefore an important part of digital accessibility – and they improve communication with every user and customer. Clear web pages also offer a better user experience and are easier for search engines to understand.

Plain language and easy language: the difference in brief

There are two levels of understandable language:

• Plain language makes texts understandable for a broad audience: short sentences, common words, a clear structure. The international standard ISO 24495-1 sets out principles and guidelines for it.

• Easy language goes further. It is designed for people with learning difficulties and follows strict rules: very short sentences, one statement per sentence, explanations for difficult words and supporting images.

You can find a detailed comparison in our article on plain language and easy language.

What makes a text easy to understand?

Plain language guidelines around the world share the same core principles. Whether a person or an AI writes the text, these plain language criteria apply:

• Put the most important information first: Start with what readers need to know, then add details (inverted pyramid).

• Use short sentences and active voice: "We will send you the invoice" is clearer than "The invoice will be sent to you".

• Avoid jargon and technical language: Use everyday words, and explain technical terms when you need them.

• Explain acronyms and abbreviations: Write them out the first time, or avoid them altogether.

• Keep terminology consistent: Use the same word for the same thing throughout the text.

• Structure the text: Clear headings and headlines, bullet points and plenty of white space help readers find their way.

• Use clear formats: Write dates and times unambiguously, for example "23 October 2026, 2 pm". In multilingual content, watch out for false friends.

A clear structure also helps people with visual impairments who use assistive technologies such as screen readers: they can jump from heading to heading and find information faster.

How readability is measured

Readability tests such as the Flesch Reading Ease score or the Flesch-Kincaid grade level calculate how difficult a text is, based on sentence and word length. They are useful for a quick check, but they do not measure reading comprehension. Whether people really understand a text can only be shown by a test for understanding with real readers.

Plain language around the world

Plain language is now an international topic:

• ISO plain language standard: The International Organization for Standardization published ISO 24495-1 in 2023. Plain language experts from many countries, including the International Plain Language Federation, contributed to it.

• United States: The Plain Writing Act of 2010 requires federal agencies to communicate in plain language. The Center for Plain Language promotes clear communication.

• European Union: The European Commission's translation service (DG Translation) publishes guidance on clear writing for EU texts.

• United Kingdom: The Plain English Campaign has campaigned for clear public information since 1979.

• WCAG: Success criterion 3.1.5 (Reading Level) recommends offering simpler content when a text requires advanced reading skills.

In content design and user-centred design, plain language has become standard practice. Content designers and digital teams write for the user, not for the organisation.

What AI can do

Modern AI-based tools have become very good at simplifying texts. They can:

• Simplify quickly: shorten long sentences, replace jargon, switch to active voice and add headings – in seconds rather than hours.

• Handle large amounts of content: Websites with hundreds of pages, blog posts, product catalogues or knowledge bases can hardly be rewritten by hand. AI makes this possible.

• Stay up to date: When content changes, AI can simplify the new version automatically.

• Work consistently: AI applies the same principles to every text, which helps with consistency across a website.

• Provide a first draft: Even when a text needs careful editing, an AI draft saves a lot of time.

For plain language, AI is therefore a very useful tool – especially for websites that want to make large amounts of content understandable.

Where people are still needed

At the same time, AI does not know everything a good text needs. People remain essential in several places:

• Context and target group: AI does not know who exactly will read a text and what they already know. A text for patients needs different explanations than a text for specialists.

• Accuracy: In legal language, medical or financial texts, simplifying must not change the meaning. Someone with expert knowledge should check whether the simplified version is still correct.

• Tone and brand: Every organisation has its own voice and its own terms. These need to be defined and checked.

• Easy language: For easy language, review by the target group is an essential part of the process. The Netzwerk Leichte Sprache states that only people with learning difficulties can say whether a text is easy enough. AI cannot replace this test.

• Images and layout: Easy language relies on supporting images and a clear layout. Choosing them requires human judgement.

• Responsibility: In the end, a person is responsible for what is published – not the AI.

That is why understandable language does not work "at the touch of a button" alone. It works best when AI and people work together.

A good workflow: AI and people together

A workflow that combines the strengths of both can look like this:

1. Define the target group and level: Should the text be in plain language for a broad audience, or in easy language for people with learning difficulties?

2. Create a draft with AI: The AI simplifies the original text and provides a first version.

3. Review and edit: An editor checks the draft for accuracy, terms and tone and adjusts it where needed.

4. Test with the target group: For easy language, people with learning difficulties check whether the text is really understandable.

5. Publish and maintain: When the original text changes, the simplified version is updated and checked again.

How Eye-Able Plain Language supports this workflow

Eye-Able Plain Language is designed for exactly this kind of workflow. Visitors can switch to a simplified version of a website with a single click. The texts are generated automatically, and when content changes, only the changed sections are processed again.

At the same time, you stay in control: in an integrated editor, you can review and adjust the simplified texts before they go live, and the module can first be tested on a preview page. This way, AI does the time-consuming groundwork, and people make sure the result fits.

Conclusion

AI has made plain language much faster and more scalable. It is an excellent tool for making large amounts of content understandable. But good, understandable texts still need people: to check accuracy, define the right tone and, for easy language, to test with the target group. The best results come when AI and people work together.

FAQ: plain language and AI

Can AI write plain language?

Yes. AI can simplify texts very well and quickly, for example by shortening sentences and replacing technical terms. For important texts, a person should still check the result for accuracy and tone.

Can AI write easy language?

AI can create a first draft in easy language. However, easy language also requires a review by people with learning difficulties. Only they can judge whether a text is easy enough. AI cannot replace this test.

How do I know if a text is easy to understand?

Readability tests such as the Flesch Reading Ease score give a first indication. They measure sentence and word length, but not whether people really understand the text. The most reliable method is a test for understanding with real readers from your target group.

What is the difference between plain language and easy language?

Plain language makes texts understandable for a broad audience. Easy language follows stricter rules and is designed for people with learning difficulties, for example with very short sentences, explanations of difficult words and supporting images.

Does AI-generated plain language replace human editing?

No. AI saves a lot of time and is ideal for large amounts of content. But people should check legal, medical or financial texts in particular, because simplifying must not change the meaning.

How does Eye-Able Plain Language work?

Eye-Able Plain Language is added to a website with a single line of code. Visitors can switch to a simplified version with one click. Editors can review and adjust the simplified texts in an integrated editor before they go live.