AI Security Communities

Independent blog with deep technical write-ups on prompt injection mechanics, agent security flaws, and LLM trust boundaries.

GitHub list aggregating 200+ LLM security tools, papers, datasets, and attack write-ups.

DEF CON community running AI red teaming events, CTFs, and publishing adversarial ML research.

Mozilla's GenAI bug bounty program rewarding AI vulnerability discovery and responsible disclosure.

Dataset of 15,000+ real-world LLM jailbreak prompts collected from Reddit and Discord communities.

Catalog of known prompt injection defense strategies with effectiveness notes and implementation links.

Repository of copy-paste LLM attack payloads organized by technique: injection, jailbreak, exfiltration.

GitHub list of MCP protocol security tools, vulnerability write-ups, and scanning resources.

GitHub list focused on autonomous agent attack surfaces: tool abuse, privilege escalation, and multi-agent risks.

OWASP working group producing checklists and deployment guidance for generative AI security.

Publishes write-ups on model serialization attacks, agent exploitation, and novel AI attack techniques.

Research portal with MITRE ATLAS-mapped detection rules and AI threat analysis.

Publishes vulnerability disclosures, tool releases, and analysis of ML supply chain risks.

Maps IAM privilege escalation paths across AWS, Azure, and GCP AI/ML services.

Live tracker of zero-day vulnerabilities in cloud infrastructure used by AI platforms.

Detection engineering blog with coverage of AI-augmented attack techniques and defense strategies.

Link collection from Arcanum covering offensive AI tools, defensive frameworks, and training materials.

Blog on prompt versioning, A/B testing, and observability patterns for LLM applications.