Terms of Service (ToS)
Effective Date: July 1, 2026 • Version 2.4 (Enterprise AI Release)
1. Agreement to Terms & Platform Overview
By creating an account, embedding the ochreshift conversational widget, or upgrading to a paid subscription, you enter into a legally binding agreement with ochreshift AI Ltd. ("ochreshift", "we", or "our"). These Terms govern your usage of our multi-model generative artificial intelligence chatbots, real-time Helpdesk takeover dashboards, and customer data synchronization pipelines.
2. Acceptable Use Policy (AUP) & Scraping Rules
To maintain high-reliability network architectures and prevent abuse, subscribers strictly agree not to deploy ochreshift for:
- Unsolicited Automation & Spam: Deploying bots to generate bulk unverified messaging, intrusive promotional campaigns, or socially engineered social phishing schemes.
- Abusive RAG Indexing: Uploading malware, copyright-infringing commercial catalogs, or utilizing internal recursive scrapers against third-party web targets without documented administrative permission.
- Malicious Override Exploits: Attempting adversarial prompt injection attacks designed to extract underlying model weights, system prompts, or private multi-tenant memory stores.
- Illegal & Deceptive Trade: Promoting illegal gambling, unlicensed controlled therapeutics, synthetic non-consensual media, or abusive impersonation workflows.
3. Enterprise AI Data Sovereignty Guarantee
We recognize that commercial business knowledge base files contain proprietary corporate secrets and customer intelligence. Under our strict zero-retention Enterprise Data Sovereignty pledges:
4. High-Risk Medical, Legal & Financial Exemption
ochreshift automated conversational engines are engineered for commercial customer support, general service FAQ resolution, and appointment onboarding. Subscribers explicitly acknowledge and agree to strict high-risk operational domain exclusions:
- Medical Diagnosis & Triage: ochreshift cannot be deployed as an autonomous clinical healthcare counselor, diagnostic device, or emergency medical dispatch engine.
- Legal & Regulatory Representation: Responses generated by ochreshift agents do not constitute formal legal representation, statutory defense, or attorney-client privileged interaction.
- Automated Credit & Employment Underwriting: Prohibiting sole reliance on AI inference for loan approvals, real estate deed closings, automated recruiting disqualification, or actuarial insurance pricing.
5. Limitation of Liability & Generative AI Disclaimers
While ochreshift incorporates grounded Retrieval-Augmented Generation (RAG) algorithms, multi-model failover chains, and adversarial guardrails, large language models natively possess statistical variance that can result in inaccurate outputs ("hallucinations").
To the maximum extent permitted by governing corporate law, ochreshift disclaims liability for commercial losses, missed customer appointments, contractual inaccuracies, or revenue variance resulting from automated agent responses. Subscribers assume primary obligation for reviewing chat analytics and configuring fallback live human intervention thresholds.
6. Subscription Termination & Default Rules
Platform administrators reserve the unilateral right to suspend AI inference endpoints or terminate user accounts immediately without financial reimbursement upon verification of deliberate Terms of Service violations, payment dispute chargebacks, or continuous automated abusive bandwidth ingestion.
Upon formal voluntary subscription termination or account deletion through the Studio security portal, all customer ledgers, custom embeddings, and lead rosters are permanently scrubbed from active storage within 30 days pursuant to our Privacy & Data Governance Policy.