AI Terms of Use

Last updated: April 7, 2026

1. AI-Generated Content

Ocean AI uses multiple artificial intelligence models (from Anthropic, Google, and OpenAI) to generate responses. These responses are:

• Generated based on patterns in training data and our proprietary knowledge base
• Not guaranteed to be accurate, complete, or current
• Provided with inline citations where possible, but citations should be independently verified
• Not a substitute for professional advice (legal, financial, tax, medical)

2. Multi-Layer Processing

Your questions are processed through a multi-layer AI pipeline:

• Layer 1 (Retrieval): An AI searches our knowledge base for relevant information
• Layer 2 (Analysis): An AI evaluates and filters the retrieved information
• Layer 3 (Synthesis): An AI generates a final response with citations

Different AI models may be used at each layer. We may run A/B tests to compare model performance.

3. Data Processing

• Your questions are sent to third-party AI providers for processing
• Your company profile context may be included in AI prompts to personalize responses
• AI providers may retain your data according to their own policies
• We log all conversations for quality assurance and abuse prevention

4. Feedback and Model Improvement

When you provide thumbs up/down feedback on responses, this data is used to:

• Evaluate model performance through ELO scoring
• Determine which model configurations produce better results
• Improve future response quality

5. Limitations

• AI may produce inaccurate or "hallucinated" information
• Citations reference our knowledge base but may not reflect the most current information
• The AI does not have access to real-time market data unless sourced from curated web sources
• Response quality depends on the clarity and specificity of your question

6. Anti-Distillation

Systematic extraction of knowledge base content through repeated queries is prohibited. We monitor for and may restrict accounts exhibiting extraction behavior.