Split – AI Safety in Education: A Teacher’s Guide to Risks, Privacy, Wellbeing and Responsible Use

Location: Split, Croatia

Course Description

AI is already in your students’ pockets. Tools like ChatGPT, Claude, Gemini, DeepSeek, Copilot and Snapchat’s My AI are used daily by young people, often without any guidance, any rules or any understanding of what happens to the information they type in. Many schools are still deciding whether to “allow” or “ban” AI, while students are already using it at home, on their phones and inside the apps they love.

This course takes a different starting point: safety first. Not fear, not panic, and not a ban. Over one week in the historic and sun-drenched city of Split, you will build a clear, practical picture of every major risk AI brings to the school environment and, more importantly, the habits, policies and classroom routines that keep students, colleagues and institutions safe.

Through hands-on workshops, real case studies and group policy design, you will leave with the confidence to lead the AI conversation in your school, rather than react to it. And with the relaxed Dalmatian “Pomalo” spirit, you will learn to do it calmly, one step at a time.

KNOW THE RISKS: THE FULL MAP

Most teacher training covers one or two AI risks, usually cheating and privacy. This course covers the whole landscape, so nothing catches you off guard. We explore:

• Data privacy and GDPR: What happens to student names, grades, photos, voices and essays once they are pasted into an AI tool? Which tools are safe for school use, and which are not?
• Hallucinations and misinformation: Why AI confidently invents facts, sources and quotes, and how errors spread through lesson plans, worksheets and student work.
• Bias and discrimination: How algorithms reproduce stereotypes about gender, ethnicity, language, disability and social background, and what that means for grading, feedback and learning materials.
• Deepfakes and synthetic media: Fake images, cloned voices and manipulated video, including their use for bullying, humiliation and impersonation of teachers and students.
• Cyberbullying and harassment: How AI makes abuse faster, cheaper and harder to trace, and how schools can respond.
• Emotional dependence and AI companions: Chatbots designed to feel like friends or partners, and what they mean for lonely, anxious and vulnerable young people.
• Harmful content and unsafe advice: What AI may say about self-harm, eating disorders, drugs, extremism or risky behaviour, and how to recognise warning signs in students.
• Cybersecurity and scams: AI-powered phishing, voice-cloning fraud, malicious “free” AI apps, prompt injection and risky browser extensions and AI agents.
• Academic integrity and cognitive offloading: Cheating, over-reliance and the quiet erosion of memory, writing and reasoning skills.
• The unreliability of AI detectors: Why detection software produces false accusations, and why it is especially unfair to non-native speakers and neurodivergent students.
• Copyright and intellectual property: Who owns AI-generated work, what training data means for authors and artists, and what students and teachers may legally publish.
• Equity and the digital divide: Who gets access to powerful tools, who does not, and how AI can widen gaps between students and schools.
• Professional risks for teachers: Deskilling, blind trust in automated output, over-surveillance, vendor lock-in and unclear legal responsibility when AI is used in decisions about children.

PROTECT YOUR STUDENTS AND YOUR SCHOOL

The safest school is not the one that bans AI. It is the one that knows what it is doing. We show you how to run a simple risk assessment for any AI tool before it enters your classroom: what data it collects, where that data goes, what age it is meant for, who is responsible if something goes wrong, and what the alternatives are.

You will practise reading terms of service and privacy policies in plain language, comparing approved and unapproved tools, and anonymising student information so that AI can still be useful without exposing anyone. We also cover the EU legal framework, including GDPR principles and the EU AI Act, in clear, teacher-friendly terms.

STUDENT WELLBEING IN AN AI WORLD

AI safety is not only about data and cheating. It is about children and teenagers. This module looks at the human side of the technology: AI companions and chatbot friendships, harmful or sexualised synthetic images, pressure from perfect-looking AI content, and the quiet isolation that can follow when a screen replaces a conversation.

You will learn how to spot early warning signs, how to talk to students without judgement, how to involve parents and school counsellors, and how to build a classroom culture where young people feel safe asking for help when something online goes wrong.

TRUTH, TRUST AND CRITICAL THINKING

In a world where text, images, audio and video can all be generated in seconds, “seeing is believing” no longer works. We teach you practical verification routines you can use yourself and pass on to students: lateral reading, source tracing, reverse image search, spotting AI-generated media, and asking the question that matters most: “How do we know this is true?”

We also explore how to protect the learning loop. When AI removes the struggle of learning, students can feel confident without being competent. You will design “Interruption Points” (short classroom routines that bring reasoning, explanation and accountability back into AI-supported tasks).

LEARNING OUTCOMES

• The Full Risk Map: Identify and explain the main categories of AI risk in education, from privacy and bias to deepfakes, scams and emotional dependence.
• Safe Tool Selection: Use a simple checklist to evaluate any AI tool for data protection, age-appropriateness, transparency and reliability before using it with students.
• GDPR and the EU AI Act in Practice: Understand your legal and ethical duties as a teacher when handling student data and using AI in a school context.
• Protecting Students: Respond step by step when AI-generated content is used to harass or impersonate someone, and recognise signs that a student is over-reliant on AI companions or exposed to harmful advice.
• Integrity and Verification: Understand why AI detectors are unreliable, design fair assignments, and teach students to fact-check AI output and spot synthetic media.
• Your Own Digital Safety: Protect your accounts, devices and professional identity from AI-powered phishing, voice cloning and malicious apps.
• The Educator’s Safety Toolkit: Leave with ready-to-use risk checklists, parent templates, student agreements, incident flowcharts and policy drafts, plus the “Pomalo” mindset to lead calmly instead of reacting anxiously.

THE COURSE FEE INCLUDES

• Free Split walking tour – with the local expert guide (included);
• Full programme of study sessions ( minimum 25 hours per week );
• Course materials;
• Certificate of attendance;
• Europass, on request
• Learning Agreement, on request
• Corporate/discount accommodation prices ;
• Visits to local institutions;
• Personalized support;
• Expertise of Local Provider on location during the mobility;
• Free WIFI internet;
• Coffee and tea breaks;
• Administration costs.

Work on your professional and personal development under the Erasmus+ network, build strong relationships with your colleagues from all over Europe, and explore amazing culture and nature in Croatia’s beautiful city of Split!

📍 Location: Erasmus Courses Croatia Split 

Explore all confirmed courses HERE