AI Use Policy

AskBiz uses Claude by Anthropic as its AI reasoning engine. Claude is a large language model (LLM) developed by Anthropic PBC, incorporated in San Francisco, California.

When you ask a question in AskBiz, the following happens:

1. AskBiz retrieves relevant data from your connected sources

2. That data, along with your question, is formatted into a prompt

3. The prompt is sent to Anthropic's Claude API via an encrypted HTTPS connection

4. Claude generates a response, which is returned to AskBiz

5. AskBiz presents the response in the interface

The data transmitted to Anthropic is transient — it is processed to generate your response and not retained by Anthropic beyond the duration of the API call. Anthropic's API terms prohibit using API requests to train models.

The following data may be included in prompts sent to Anthropic's Claude API:

• Relevant excerpts of your business data (sales figures, product names, revenue metrics) necessary to answer your question
• The text of your question
• Contextual information about your business type, sector, and settings (e.g. home currency, connected platforms)

The following is never sent to Anthropic:

• Full datasets from your connected sources (only relevant excerpts)
• Customer personal data (email addresses, names) — these are hashed in our system before any processing
• Your payment details or billing information
• Credentials or OAuth tokens for connected platforms
• Other users' data

Under the EU AI Act (Regulation 2024/1689), AI systems are classified by risk level. AskBiz is classified as a limited-risk AI system (not high-risk) because:

• It provides analysis and recommendations to support human business decisions
• It does not make autonomous decisions about individuals with significant legal or economic effects
• It does not operate in high-risk sectors listed in Annex III (critical infrastructure, employment decisions, access to essential services, law enforcement, etc.)
• Human review and override is always possible and expected

AskBiz complies with the transparency obligations applicable to limited-risk AI systems under Article 50 of the EU AI Act.

AI outputs are generated probabilistically and may contain errors. AskBiz:

• Displays confidence indicators on all AI-generated responses
• Encourages users to verify important figures against source data
• Does not guarantee the accuracy of AI-generated analysis
• Is not liable for business decisions made solely in reliance on AI-generated outputs without independent verification

This does not limit your statutory rights under UK or EU consumer protection law where applicable. AskBiz's liability limitations are set out in the Terms of Service.