Inqyra WP plugin
Download
FeaturesPricingHow It WorksSupportDocumentationBlog Download

1. Nature of AI Responses

1.1 AI-Generated Content

The Inqyra plugin generates chatbot responses using third-party AI language models. Responses are produced through Retrieval-Augmented Generation (RAG): the plugin retrieves relevant content from the Customer’s knowledge base and sends it as context to the selected AI provider, which generates a response.

1.2 No Human Review

AI responses are generated automatically without human review or approval. Neither Ipportunities B.V. (“Ipportunities”) nor the Customer’s website necessarily reviews individual responses before they are displayed to Visitors.

1.3 Not Ipportunities’ Content

Ipportunities provides the software infrastructure. The content of AI responses is generated by third-party AI models based on the Customer’s knowledge base and configuration. Ipportunities does not control, curate, or endorse the content of AI-generated responses.

2. Multi-Provider Architecture

2.1 Customer-Selected Provider

The Customer chooses which AI provider to use and provides their own API key. The plugin currently supports:

Provider Terms of Service
Anthropic (Claude) https://www.anthropic.com/policies
OpenAI https://openai.com/policies
Google (Gemini) https://ai.google.dev/terms
Mistral https://mistral.ai/terms
DeepSeek https://www.deepseek.com/terms

Note: URLs above are provided for reference and may change. Verify current links before publication. [PLACEHOLDER: verify all URLs]

2.2 Direct Communication

API calls are made directly from the Customer’s server to the AI provider. Ipportunities does not proxy, store, or have access to these communications. The Customer’s relationship with the AI provider is governed by the provider’s terms of service.

2.3 Provider Responsibility

Each AI provider is independently responsible for its model’s behavior, content policies, and data handling. Ipportunities has no control over how providers train their models, filter outputs, or process data.

3. Accuracy Limitations

3.1 No Guarantee of Accuracy

AI language models may produce responses that are:

  • Incorrect: Factually wrong or containing fabricated information (“hallucinations”);
  • Incomplete: Missing important context, caveats, or details;
  • Outdated: Based on training data that does not reflect current information;
  • Misleading: Appearing confident while being wrong;
  • Inconsistent: Providing different answers to the same question across sessions.

3.2 Knowledge Base Dependency

The quality of AI responses depends heavily on the Customer’s knowledge base. The plugin retrieves relevant text chunks (up to 1,000 characters each) and provides them as context to the AI model. Responses are only as good as the source material.

3.3 Not a Substitute for Professional Advice

AI-generated responses do NOT constitute:

  • Legal advice — Consult a qualified lawyer;
  • Medical advice — Consult a qualified healthcare professional;
  • Financial advice — Consult a qualified financial advisor;
  • Tax advice — Consult a qualified tax professional;
  • Emergency guidance — Contact the appropriate emergency services.

AI responses should not be relied upon for decisions that could have significant consequences for health, safety, finances, or legal rights.

4. Content Responsibility

4.1 Ipportunities

Ipportunities provides the software platform that enables the chatbot. Ipportunities is responsible for:

  • The software functioning as described;
  • License Server operation;
  • Security of the software code.

Ipportunities is NOT responsible for:

  • The content of AI-generated responses;
  • The accuracy or completeness of the Customer’s knowledge base;
  • The Customer’s choice of AI provider;
  • How Visitors interpret or act on AI responses.

4.2 Customer (Website Operator)

The Customer is responsible for:

  • The content of their knowledge base (crawled pages, uploaded documents);
  • Configuring the AI system prompt and guardrails;
  • Selecting an appropriate AI provider for their use case;
  • Monitoring chatbot conversations and quality;
  • Informing Visitors that they are interacting with an AI chatbot;
  • Compliance with applicable laws regarding AI-generated content.

4.3 AI Provider

The AI provider is responsible for:

  • The behavior and capabilities of its model;
  • Content filtering and safety measures;
  • Data processing in accordance with its own terms and privacy policy.

5. Lead Extraction (Premium Feature)

5.1 Automated Processing

The premium lead extraction feature automatically identifies and extracts personal data from chat conversations, including email addresses, phone numbers, names, and company names. This constitutes automated processing of personal data.

5.2 Disclosure Obligation

Customers who enable lead extraction should:

  • Inform Visitors that personal data may be automatically extracted from conversations;
  • Include this processing in their privacy policy;
  • Ensure a valid legal basis exists for this processing (e.g., legitimate interest, consent);
  • Assess whether a DPIA is required under Art. 35 GDPR.

5.3 Data Storage

Extracted lead data is stored exclusively in the Customer’s WordPress database. Ipportunities does not have access to this data.

6. DeepSeek — Special Considerations

6.1 Data Processing in China

DeepSeek is headquartered in China and processes data on servers located in China. The People’s Republic of China is not recognized by the European Commission as providing an adequate level of data protection under Art. 45 GDPR.

6.2 Legal Implications

Customers located in the EEA who use DeepSeek should be aware that:

  • Chinese national security and intelligence laws may require DeepSeek to provide access to data upon government request;
  • Standard Contractual Clauses (SCCs) may be insufficient without supplementary measures;
  • A Transfer Impact Assessment (TIA) is recommended before use;
  • Supervisory authorities in the EU have expressed concerns about data transfers to China.

6.3 Customer Responsibility

The decision to use DeepSeek is made entirely by the Customer. Ipportunities provides the technical integration but does not recommend or endorse DeepSeek for use cases involving personal data of EU data subjects. Customers should independently assess the risks and legal basis for this data transfer.

7. Embeddings

7.1 Browser-Based Processing

Text embeddings (vector representations used for semantic search) are generated entirely in the browser using the Xenova/paraphrase-multilingual-MiniLM-L12-v2 model via the Transformers.js library. No content is sent to any external API for embedding generation.

7.2 No External Data Transfer

The embedding process does not involve any data transfer to Ipportunities or any third-party service. The model runs locally in the browser, and resulting vectors (384 dimensions) are stored in the Customer’s database.

8. Liability

8.1 No Liability for AI Output

To the maximum extent permitted by applicable law, Ipportunities shall not be liable for any claims, damages, losses, or expenses arising from:

  • The content, accuracy, or completeness of AI-generated responses;
  • Decisions made or actions taken based on AI-generated responses;
  • Any harm resulting from reliance on AI-generated information;
  • Content generated by Third-Party AI Providers.

8.2 Overall Limitation

Liability related to the Software, including AI functionality, is governed by the limitation of liability provisions in the EULA (Section 11).

9. Recommendations for Website Operators

Customers deploying the Inqyra chatbot should consider the following best practices:

9.1 Visitor Disclosure

  • Display a visible notice that the chatbot is AI-powered;
  • Inform Visitors that responses may not always be accurate;
  • Provide a way for Visitors to contact a human for critical inquiries.

9.2 Configuration

  • Write a clear system prompt with appropriate guardrails;
  • Instruct the AI to stay within the scope of the knowledge base;
  • Configure the AI to disclose its limitations when appropriate;
  • Test the chatbot thoroughly before deployment.

9.3 Monitoring

  • Regularly review conversation logs for quality and accuracy;
  • Keep the knowledge base up to date;
  • Monitor for inappropriate or harmful responses;
  • Have a process for quickly disabling the chatbot if issues arise.

9.4 Legal Compliance

  • Include AI chatbot usage in your privacy policy;
  • Assess whether the AI Regulation (EU AI Act) applies to your use case;
  • If operating in a regulated industry (healthcare, finance, legal), obtain professional guidance on AI use;
  • Document your AI risk assessment.

10. Changes

This AI Disclaimer may be updated to reflect changes in supported providers, features, or regulatory requirements. Material changes will be communicated via the Software’s admin interface.

Ipportunities B.V.

KvK: [PLACEHOLDER]

Address: [PLACEHOLDER]

Email: [PLACEHOLDER]

Related documents: EULA · Privacy Policy · Data Processing Agreement · Support Policy