Prompt Injection: Hidden Security Risk in AI | AI Security Webinar

About this event

Prompt Injection: The Hidden Security Risk in AI Prompt injection is one of the most important security risks facing AI applications, LLMs and AI agents today. As businesses increasingly build with ChatGPT, large language models, AI assistants, RAG systems and AI agents, understanding how untrusted instructions can influence an AI system is becoming an essential AI security skill. Join the AI Learners Club for a beginner-friendly online webinar exploring prompt injection, LLM security, AI application security and practical ways to build safer AI systems. No cybersecurity or AI engineering experience is required. 📅 Date: 15 October 2026 ⏰ Time: 6:00 PM Sydney time 💻 Location: Online 🎓 Level: Beginner friendly What is Prompt Injection? Prompt injection occurs when untrusted input changes an AI system's behaviour in a way the application did not intend. This can happen directly when a user attempts to override an AI application's instructions, or indirectly when malicious instructions appear inside content the AI is processing, such as a webpage, document, email or file. During this webinar, we'll break the concept down without unnecessary jargon and explore why prompt injection becomes particularly important when AI systems can access external tools, sensitive information or automated actions. What You'll Learn 🔹 What prompt injection is and why it matters 🔹 The difference between direct and indirect prompt injection 🔹 How malicious or untrusted content can influence an LLM 🔹 What prompts can — and cannot — actually do 🔹 Why AI tool access and excessive permissions increase security risks 🔹 Why AI-generated output should be treated as untrusted 🔹 How to separate trusted instructions from untrusted data 🔹 How least privilege can reduce the impact of AI security failures 🔹 Practical patterns for validating AI output before taking actions 🔹 How developers can design AI applications so model mistakes don't become security incidents We'll also walk through a safe prompt injection demonstration, practical scenarios and an interactive discussion where participants can identify vulnerabilities and think through potential protections. The three core security principles we'll explore are: 1\. Don't automatically trust AI output Validate AI responses before allowing them to trigger important actions. 2\. Separate instructions from data User input, documents, websites and external content should be treated as untrusted data rather than application instructions. 3\. Limit permissions Give AI systems access only to the tools, information and actions they actually need. Who Should Attend? This webinar is suitable for: • AI beginners • Developers and software engineers • Cybersecurity professionals and students • AI enthusiasts • Product managers and founders • People building with ChatGPT or LLM APIs • AI agent and automation builders • Anyone interested in AI safety, AI security or responsible AI You don't need to know how to code to participate. Speakers Saanvi Nayak AI Security Enthusiast & Community Builder https://www.linkedin.com/in/saanvi-nayak-1926b0403/ Dev Roy President, AI Learners Club https://www.linkedin.com/in/aihistorian/ https://ailearnersclub.com/