Blog & Resources
Security awareness tips, phishing prevention guides, and best practices.
Claude Mythos 5 Now in Claude Security
Claude Mythos 5 is now available in Claude Security, with partner availability coming soon to enhance cyber defense capabilities.
Anthropic launches $35M fund for open-source cyber defense
Anthropic announced a $35 million fund to support open-source cybersecurity projects, aiming to democratize access to advanced AI tools for global def
OWASP Unveils Top 10 AI Security Risks
OWASP released a new top 10 security list for AI, introducing a Universal Skill Format to standardize and secure AI add-ons.
AI Generates Viable Bacteriophages
AI models designed 285 bacteriophage genomes, and 16 proved viable, attacking E. coli.
Wazuh and AI For Enhanced SOC Workflows
AI is transforming cybersecurity by automating tasks and uncovering patterns.
AI Passwords Are Predictable, Hackers Can Guess
AI-generated passwords use repetitive patterns that hackers can exploit, making them easier to guess than expected.
AI Attacks at Machine Speed, CEO Says
AI-powered cyberattacks now operate at machine speed, fundamentally changing the cybersecurity landscape and investment case, according to a CEO.
Encrypted Prompts Bypass AI Safety in Grok, Gemini
Researchers found 'Cryptographic Context Injection' hides malicious instructions in encrypted prompts, bypassing safety guardrails in Grok and Gemini.
Korea Updates AI Ethical Principles After Five Years
South Korea has revised its AI ethical principles for the first time in five years, addressing evolving concerns about AI-related harms and governance
Eleven Quick Tips for Biomedical Federated Learning
Federated learning enables collaborative biomedical research without sharing raw data, but requires careful design.
OpenAI's training pause sets safety precedent
OpenAI's training pause establishes a safety precedent, but only leading AI firms can afford such measures.
OpenAI Adds AI Security Controls After Escape
OpenAI introduced new AI security controls following last month's Hugging Face incident, though critics note these should have existed earlier.