Topic: Prompt Injection: Understanding AI Security Vulnerabilities
Abstract: Large Language Models (LLMs) are rapidly transforming business operations, software development, customer engagement, and cybersecurity workflows. However, as organizations increasingly integrate AI-powered applications into their environments, new attack surfaces and security risks are emerging. Threat actors are actively exploring ways to manipulate, exploit, and compromise LLM-based systems through techniques such as prompt injection, data poisoning, model manipulation, sensitive data extraction, and supply chain attacks.
The current webinar examines the evolving security landscape surrounding LLMs and generative AI applications. The session explores common attack vectors, real-world threat scenarios, and emerging security frameworks designed to secure AI deployments. Join us to gain insights into how AI systems are attacked, the vulnerabilities that exist across the AI lifecycle, and the potential business impact of successful exploitation.
Key Takeaways:
- Understanding the security architecture and attack surface of modern LLMs
- How prompt injection attacks manipulate AI behavior and bypass safeguards
- Risks associated with RAG, AI agents, plugins, and third-party integrations
- Understanding the exploitation of AI supply chains and model dependencies
- Best practices for securing AI applications throughout the development lifecycle
- Emerging AI security frameworks, standards, and governance requirements
Speaker:
Céline Blandin, Cybersecurity Professional and AI Security Researcher.
Bio: Céline Blandin is a cybersecurity and AI professional focused on making emerging AI risks understandable to everyone. With experience in penetration testing and AI system evaluation, she explores how tools such as large language models can introduce new, sometimes unexpected, security challenges. She has contributed to AI-related projects and regularly shares her work and research online. Céline has delivered talks on cybersecurity and AI to a wide range of audiences, from students to professionals, with an emphasis on clarity and real-world examples. Her goal is to help people better understand how AI works and how to use it more safely.
