Brent Coker: AI security guides and insightsImplementation-level analysis of AI security problems, grounded in the research literature: prompt injection, RAG isolation, agent tool use and red-teaming.https://brentcoker.com/feed.xml2026-11-10T00:00:00ZDr Brent Cokerhttps://brentcoker.com/https://brentcoker.com/favicon.svgSecuring Autonomous AI Agents: Deterministic Gateways for Tool Use and Function Callinghttps://brentcoker.com/insights/securing-ai-agent-tool-calling2026-11-10T00:00:00Z2026-11-10T00:00:00ZAI agent security fails when agents hold their own credentials. A policy gateway, strict tool-call schemas and rare, clear approvals contain excessive agency.Mitigating Indirect Prompt Injection in Multi-Tenant RAG Pipelineshttps://brentcoker.com/insights/indirect-prompt-injection-rag2026-11-10T00:00:00Z2026-11-10T00:00:00ZHow untrusted documents, emails and web pages hijack RAG sessions, and the privilege boundaries and deterministic output checks that contain the damage.