EU AI Act transparency

Limited-risk transparency intent (Art. 50) — product starting package.

Under legal review: this page is a preview draft that is currently being reviewed by counsel and may change. Status: 2026-10-26.

Sentinel uses automated monitoring, triage, and (on Guardian+) optional auto-fix workflows. You are interacting with an AI software employee, not a human livestream.

Simulated employee presence (desk portrait, LIVE badge, captions, optional browser speech) is a UI stub — not a real human on camera and not proof of continuous human oversight.

Action logging: incident outcomes, demo timelines, KYC ops notes, and EU AI Act audit log (not a SOC 2 certification) JSONL events are written under the app data/ directory for transparency and debugging.

Provider: WashedBlackDryTrading Product & Software Engineering, Köln.

Human oversight: turn off auto-fix, re-run demos, erase customer data via DELETE /customer/{id}, stop the service. Residual risk: models and heuristics can be wrong; do not rely on auto-fix alone for critical production systems.