Uncovering DeepSeek: Key Findings from the Qualys TotalAI Security Analysis - 18 de marzo de 2025 - TecnoWebinars.com

Generative AI tools are rapidly transforming industries, yet the security risks of deploying Large Language Models (LLMs) remain widely underexplored. In this live webinar, join Rijul Nirwal, Product Manager for Qualys TotalAI, and Colton Pepper at Qualys, as they reveal findings from our recent security assessment of DeepSeek-R1 using Qualys TotalAI. We’ll showcase how adversarial attacks, prompt vulnerabilities, and misalignment issues can affect not just DeepSeek, but any LLM platform. By balancing both technical and strategic perspectives, we’ll help cybersecurity professionals, AI/ML developers, and C-level leaders understand the real-world implications of these vulnerabilities and why they matter for business. Key takeaways include:

• Adversarial Threats: See how common jailbreak and prompt injection attacks can expose LLM-based solutions to misuse.
• Vulnerability Highlights: Understand the categories of misalignment issues, data privacy pitfalls, and potential operational risks.
• Governance and Oversight: Learn why strong AI oversight and compliance considerations are critical in enterprise settings.