DGB 4: AI – Opportunities and Limitations

← Digital Literacy – Year 4

Artificial intelligence is already used in search, translation, navigation, image recognition, recommendation systems, accessibility tools and generative applications. Digital competence in Year 4 therefore means more than being able to use an AI tool: students should understand what such systems can do, where their limits lie, how errors arise and why people remain responsible for checking important results.

Curriculum reference

Area	Classification
Competence area	Orientation
Perspectives	T, G, I
Focus	Reflecting on the opportunities and limitations of artificial intelligence
Competence	Explain core ideas of AI and critically evaluate its capabilities, errors, social consequences and human responsibility.

Learning objectives

Students can:

explain the terms artificial intelligence, machine learning and generative AI;
describe typical applications of AI in everyday life, school and the world of work;
explain opportunities in medicine, transport, translation, accessibility and environmental monitoring;
identify technical limits involving data quality, context, rare cases and model errors;
recognise bias, black-box effects and confidently worded false statements;
verify AI output with sources, subject knowledge and human judgement;
discuss effects on privacy, fairness, work, creativity and democracy;
formulate rules for safe, fair and responsible use of AI.

Learning content

What artificial intelligence is

AI systems process data and use models to identify patterns, classify inputs, make predictions or generate new content. Machine learning develops patterns from examples rather than relying only on explicitly programmed rules. Generative AI produces text, images, audio or code from probability-based models. These systems do not possess human understanding simply because their output sounds natural.

Opportunities of AI

AI is particularly useful when large data sets need to be analysed quickly, repeated tasks can be automated or information needs to be adapted to users. Examples include speech recognition, automatic captions, translation, traffic prediction, image analysis, spam filtering and assistance for people with disabilities.

Technical limitations

AI quality depends on training data, model design, current context and the task. A model can miss unusual cases, reproduce incorrect patterns or generate plausible but false information. Good language is not proof of factual accuracy.

Social and ethical limitations

AI can affect privacy, access to opportunities, employment and public debate. Biased data can lead to unfair outcomes. If a system is difficult to explain, it may be unsuitable for high-stakes decisions without additional safeguards.

Human responsibility and critical use

People decide when an AI tool is appropriate, which information is entered, how output is checked and who is accountable for consequences. Responsible use includes protecting sensitive data, citing or documenting AI assistance where required and keeping human judgement in important decisions.

Questions for reflection

Useful questions are: What data might the system use? What could go wrong? Who benefits? Who might be disadvantaged? Can the result be checked? Is a person responsible for the final decision? What would happen if the system were unavailable?

Classroom activities

Check AI answers

Compare AI-generated answers with reliable sources and mark correct, uncertain and false statements.

Expected outcome: A documented fact-check.

Human versus AI

Compare a suitable task completed by a person and by an AI tool using quality, time, transparency and responsibility as criteria.

Expected outcome: A balanced comparison.

Fairness case study

Analyse a fictional case in which biased training data leads to unequal treatment.

Expected outcome: A cause-and-consequence analysis.

Debate responsibility

Discuss who is responsible when an AI-supported decision causes harm.

Expected outcome: A reasoned responsibility map.

Prompt and result quality

Change one prompt systematically and observe how specificity, context and constraints influence output.

Expected outcome: A short experiment log.

Create a school AI guide

Develop practical rules for privacy, verification, transparency and appropriate use.

Expected outcome: A concise class guideline.

Reflect and secure learning

Write three opportunities, three limitations and three rules for responsible AI use.

Expected outcome: A personal summary.

Self-check

I can explain AI, machine learning and generative AI in my own words.
I can give realistic examples of useful AI applications.
I can recognise important technical and social limitations.
I do not treat fluent AI output as automatically correct.
I can verify important results using suitable sources.
I can justify rules for responsible AI use.