Acceptable Use Policy
Last updated: 31 July 2026.
Customers must use UVOX lawfully and in accordance with the terms of their selected AI providers.
Prohibited use
- Unauthorised access, credential theft, malware delivery or attempts to bypass security controls.
- Processing data without required authority, consent or legal basis.
- Abusive traffic, deliberate service disruption, rate-limit evasion or automated account creation.
- Use that violates applicable sanctions, export restrictions, intellectual-property rights or provider policies.
- Misrepresentation of UVOX measurements, quality claims or affiliation with AI providers.
Security
Customers must protect UVOX and provider credentials, use separate production keys where possible and notify UVOX promptly of suspected compromise.
Enforcement
UVOX may investigate, restrict or suspend use reasonably believed to violate this policy, the Terms of Service or applicable law.
Zero-Risk and verification disclosure: “Zero-risk” applies only to Zero-Loss Cache Mode and means prompt content is not rewritten or deleted, provider/model are not switched, and generated responses are not substituted from a semantic answer cache. Provider-supported cache metadata may be added. Proof Lab requests are sent to the selected provider and UVOX using customer-supplied keys held in process memory. Reports exclude API keys and full prompt content by default. Savings and exclusivity claims are workload- and evidence-dependent and are not guarantees.