hristosbilis.aiAI security for regulated pharma — making LLM gateways and AI agents auditable, validated, and 21 CFR Part 11-defensible.https://hristosbilis.ai/en-usWhere pharma AI actually is — a 2026 ISPE AI Summit field readouthttps://hristosbilis.ai/writing/where-pharma-ai-actually-is/https://hristosbilis.ai/writing/where-pharma-ai-actually-is/Two days inside the room at the 2026 ISPE AI summit: where regulated-pharma AI is really deployed, where it's going, and the one control almost nobody owns yet.Sun, 12 Jul 2026 00:00:00 GMTSecurity is now a GMP requirementhttps://hristosbilis.ai/writing/security-is-now-a-gmp-requirement/https://hristosbilis.ai/writing/security-is-now-a-gmp-requirement/The 2025 Annex 11 rewrite puts pen-testing, MFA, and audit trails into GMP text — so for AI systems, the security controls and the validation controls are now the same controls.Sat, 11 Jul 2026 00:00:00 GMTHow much validation does your AI actually need? Start with Context of Usehttps://hristosbilis.ai/writing/how-much-validation-does-your-ai-need/https://hristosbilis.ai/writing/how-much-validation-does-your-ai-need/Before you validate a GxP AI system, classify it. A practical, cross-regime procedure — Context of Use, influence x consequence, Annex 22, GAMP 5, EU AI Act — that tells you how much evidence you actually owe.Fri, 10 Jul 2026 00:00:00 GMTWhat a CSV validator should look for in an LLM gatewayhttps://hristosbilis.ai/writing/what-a-csv-validator-should-look-for-in-an-llm-gateway/https://hristosbilis.ai/writing/what-a-csv-validator-should-look-for-in-an-llm-gateway/A pilot went well and the business wants it in production. The CSV team opens the architecture diagram and sees 'LLM gateway' in the middle. Here's what that means for validation — and the three failures to reproduce before signoff.Thu, 09 Jul 2026 00:00:00 GMTAn ALCOA+ audit trail for LLM callshttps://hristosbilis.ai/writing/alcoa-audit-trail-for-llm-calls/https://hristosbilis.ai/writing/alcoa-audit-trail-for-llm-calls/The smallest honest example of turning AI telemetry into 21 CFR Part 11 validation evidence — a tamper-evident, hash-chained audit trail for every model interaction.Tue, 07 Jul 2026 00:00:00 GMTAn LLM gateway from scratch, in ~150 lineshttps://hristosbilis.ai/writing/gateway-from-scratch/https://hristosbilis.ai/writing/gateway-from-scratch/What an LLM gateway actually does — a minimal, keyless FastAPI program between apps and model providers that adds five security controls one at a time, each answering a question a CSV/QA reviewer is already asking.Mon, 06 Jul 2026 00:00:00 GMT