Cisco Co-Authors Update to the NIST Adversarial Machine Learning Taxonomy
Cisco and the UK AI Security Institute partnered with NIST to release the latest update to the Adversarial Machine Learning Taxonomy.
Security for AI
Evaluating Security Risk in DeepSeek and Other Frontier Reasoning Models
The performance of DeepSeek models has made a clear impact, but are these models safe and secure? We use algorithmic AI vulnerability testing to find out.
Cisco AI Defense: Comprehensive Security for Enterprise AI Adoption
Cisco AI Defense is a single, end-to-end solution that helps your organization understand and mitigate risk on both the user and application levels.
Leveraging Hardened Cybersecurity Frameworks for AI Security
Strengthen AI security by leveraging hardened cybersecurity frameworks like CWE to mitigate vulnerabilities and enhance resilience.
Bypassing OpenAI’s Structured Outputs: Another Simple Jailbreak
Discover how researchers bypass OpenAI's structured outputs with advanced jailbreak techniques. Learn about the vulnerabilities, implications, and ways to enhance AI system security in this insightful blog post.
Bypassing Meta’s LLaMA Classifier: A Simple Jailbreak
Discover how researchers bypassed Meta's LLaMA classifier using a straightforward jailbreak method. Learn about the vulnerabilities in AI content moderation and the implications for AI security.
Introducing the Coalition for Secure AI (CoSAI)
Announcing the launch of the Coalition for Secure AI (CoSAI) to help securely build, deploy, and operate AI systems to mitigate AI-specific security risks.
Fine-Tuning LLMs Breaks Their Safety and Security Alignment
Fine-tuning large language models can compromise their safety and security, making them more vulnerable to jailbreaks and harmful outputs.
Securing the LLM Stack
Learn how to secure the LLM stack, which is essential to protecting data and preserving user trust, as well as ensuring the operational integrity, reliability, and ethical use of these powerful AI models.