1. Where automation may be used
- Crawling, rule evaluation, classification, prioritization, deduplication, correlation, scoring, and anomaly detection.
- Summaries, explanations, draft recommendations, tickets, report sections, and suggested verification steps.
- Queueing, monitoring, notifications, readback, and governed execution of an approved action through a configured connector.
2. No guarantee of correctness
Automated and AI-assisted output may be incomplete, incorrect, outdated, biased, inapplicable, or based on limited evidence. It can produce false positives and false negatives. A score, label, summary, or recommendation is not a guarantee of security, accessibility, legal compliance, search performance, availability, or business outcome.
3. Human responsibility
Customers remain responsible for authority, scope, professional judgment, applicable law, implementation, approvals, production changes, and verification. High-impact or irreversible action must receive appropriate human review, testing, backup, rollback planning, and post-change readback.
4. Customer Data and providers
Asuruas processes Customer Data only for the purposes and under the roles described in the Terms, Privacy Policy, and DPA. If an external AI provider will process Customer Personal Data on our behalf, that provider and purpose must be disclosed in the Subprocessor register before production use where required. We do not authorize an external provider to train a general-purpose model on Customer Data unless a separate written agreement and lawful instructions expressly allow it.
5. Production agents
A configured connector may execute only allowed actions within its tenant and site binding. Signed requests, idempotency, authorization, approval, rollback, and readback controls reduce risk but do not eliminate the need for a valid scope and operator review.
6. Questions and date
Questions about a particular automated feature or data flow may be sent to clancaster@lancastersolutionsllc.com.
Effective August 9, 2026.