5.1 AI systems can produce outputs that are inaccurate, incomplete, misleading, inconsistent, or inappropriate, and may appear authoritative despite being incorrect.
5.2 Voxd makes no warranties about AI Outputs (including accuracy, completeness, reliability, legality, or fitness for purpose).
5.3 The Client is responsible for deciding whether to use, publish, act on, or rely on any AI Output. The Client should apply appropriate human review where outputs could affect customers, money, legal rights, health, safety, or regulatory compliance.
5.4 AI Outputs are generated by third-party AI models and may be incorrect or inappropriate. Voxd is not responsible for losses caused by the Client’s or any end user’s use of or reliance on AI Outputs or for decisions made based on them (including where outputs are inaccurate, misleading, incomplete, offensive, or fail to follow the Client’s preferred tone or instructions). Voxd remains responsible for delivering the Services (configuration/integration) with reasonable skill and care.
5.5 Where the chatbot uses AI models provided by third parties, those providers control model behaviour, safety filters, pricing, availability and policies. Voxd is not responsible for changes or issues in third-party model behaviour or availability.
5.6 End users may try to manipulate AI behaviour (e.g., prompt injection). Voxd is not responsible for AI Outputs resulting from such attempts.