ConvoViva is a WordPress plugin that integrates a conversational assistant using OpenAI’s services and API.

When an AI feature is used, the plugin acts as a technical intermediary between the WordPress site and OpenAI. ConvoViva does not develop, host, or control the AI models used to produce responses.

By installing, configuring, or using ConvoViva, you acknowledge that you have read and accepted the warnings below.

1. Independent service from OpenAI

ConvoViva is independent software. It is not developed, sponsored, or endorsed by OpenAI and is not part of OpenAI’s official products.

The names «OpenAI,» «ChatGPT,» «GPT,» and the names of various models are the property of their respective owners. Their use in ConvoViva documentation is solely to describe compatibility and the extension’s technical operation.

2. API key and OpenAI account

To use AI features, the site operator must provide or configure valid access to the OpenAI API.

The site operator remains fully responsible for:

• their OpenAI account;
• the security and confidentiality of their API key;
• the choice of models used;
• the usage limits and quotas of their account;
• fees billed by OpenAI;
• any use made via their API key.

The purchase or use of a ConvoViva license does not include any OpenAI API credits, subscriptions, or usage plans. OpenAI fees are separate from those associated with ConvoViva.

3. Accuracy of responses

Responses are generated automatically by an AI model. They may be inaccurate, incomplete, ambiguous, outdated, contradictory, biased, or inappropriate.

An AI model may also present entirely or partially invented information as factual. This phenomenon is sometimes called «hallucination» or «fabrication.».

ConvoViva therefore does not guarantee the accuracy, reliability, legality, relevance, or suitability of the responses produced.

Any important information should be verified by a competent person before being used, published, shared with a third party, or relied upon to make a decision.

4. No professional advice

Responses generated using ConvoViva should not be considered legal, medical, financial, accounting, tax, psychological, technical, or professional advice.

The extension must not be used as the sole basis for a decision that could have significant consequences for a person’s health, safety, finances, rights, or obligations.

5. Site operator responsibility

The person or organization that installs ConvoViva on a WordPress site is responsible for its configuration, use, and oversight.

This responsibility includes, in particular:

• drafting and validating the instructions given to the AI;
• selecting the OpenAI model;
• monitoring the responses generated;
• preventing abuse and malicious uses;
• protecting personal information;
• managing conversation logs;
• managing conversations stored with OpenAI;
• managing files sent to knowledge bases;
• complying with applicable laws and regulations;
• complying with OpenAI’s terms and policies.

The site operator must also clearly inform visitors that they are communicating with an automated system using OpenAI’s services.

6. Transmission of data to OpenAI

When a visitor sends a message that requires a response from the AI, certain data is transmitted to OpenAI’s servers so the request can be processed.

This data may include:

• the visitor’s message;
• the instructions configured for the agent;
• the context of previous exchanges, when memory is enabled;
• excerpts from a knowledge base;
• certain technical information necessary to process the request.

Data transmitted to OpenAI are subject to OpenAI’s contractual terms, privacy policies, acceptable use policies, and data retention mechanisms.

7. Memoryless mode

In stateless mode, ConvoViva sends each request separately to OpenAI’s Responses API with the parameter store: false.

This parameter indicates that the response should not be stored as application state for later retrieval by the API. However, by itself it does not guarantee zero retention or immediate deletion of all data.

OpenAI may retain certain data or metadata for abuse detection, security, policy enforcement, or legal obligations, in accordance with the rules applicable to the OpenAI account in use.

8. Conversational memory mode

When memory mode is enabled, ConvoViva uses OpenAI’s conversation features to retain context across multiple messages.

Conversations may then include messages, responses, tool calls, and other elements associated with the exchange. These data may remain stored at OpenAI until deleted.

The site operator must choose this mode knowingly, inform users of this retention, and provide a procedure to manage or delete conversations when required.

9. Knowledge bases and files

ConvoViva can enable the creation of a knowledge base using OpenAI’s files and vector database features.

Documents added to a knowledge base may be uploaded to OpenAI, processed, indexed, and stored on its servers until deleted.

The site operator must ensure they have the rights and authorizations necessary to transmit these documents to OpenAI.

They must not upload confidential information, trade secrets, personal data, or copyrighted content without first assessing the consequences and obtaining required permissions.

10. Logs stored in WordPress

ConvoViva may store conversation logs locally on the WordPress server. Depending on the configuration and type of interaction, these logs may include, among other things:

• the date and time of the exchange;
• the visitor’s IP address;
• the identifier of a logged-in WordPress user;
• the agent and language used;
• the message sent by the visitor;
• the response returned by the artificial intelligence.

The site operator is responsible for determining whether this logging is appropriate, configuring a reasonable retention period, protecting access to the logs, and deleting them when no longer necessary.

Using ConvoViva does not replace the obligation to publish a privacy policy appropriate for the website. That policy must clearly explain what data is collected, why it is collected, where it is transmitted, and how long it is retained.

11. Personal and confidential information

Visitors should never submit passwords, API keys, payment card numbers, government identification numbers, or any other highly confidential information to the chatbot.

The site operator should also avoid requesting or processing sensitive information, including medical, financial, or legal information, unless the necessary safeguards, authorizations, and agreements are in place.

12. Use by minors

The site operator is responsible for determining whether the chatbot may be used by minors and, if so, for implementing the notices, restrictions, and parental consent mechanisms required by law and by OpenAI’s policies.

13. Intellectual property

An AI-generated response may resemble existing content, include trademarks, or reproduce elements protected by third-party rights.

It is the responsibility of the site operator and the user to verify they have the necessary rights before publishing, selling, reproducing, or distributing generated content.

ConvoViva does not guarantee the originality, exclusivity, or absence of intellectual property infringement in responses produced by OpenAI.

14. Availability and changes to services

ConvoViva’s operation depends in part on WordPress, site hosting, Internet connectivity, and the availability of OpenAI services.

OpenAI may change its models, pricing, limits, policies, APIs, or features without the ConvoViva developer being able to control those changes.

Interruptions, slowdowns, errors, request rejections, quota exceedances, or temporary incompatibilities may therefore occur.

15. Prohibited or high-risk uses

ConvoViva must not be used to facilitate illegal, fraudulent, deceptive, discriminatory, dangerous, or malicious activity.

The extension should not be configured to make high-risk decisions about a person on its own, including regarding employment, credit, insurance, housing, health, education, law enforcement, or access to an essential service.

Any use of the OpenAI API must comply with OpenAI’s usage policies and the laws applicable in the jurisdiction of the site operator and its users.

16. Limitation of liability

To the extent permitted by law, ConvoViva is provided «as is» and «as available,» without express or implied warranty of uninterrupted operation, accuracy of responses, or fitness for a particular purpose.

The ConvoViva developer cannot be held responsible for responses generated by OpenAI, decisions made based on those responses, use of the plugin, or data transmitted by the site operator and its visitors.

To the extent permitted by law, the developer cannot be held liable for data loss, loss of revenue, business interruption, reputational harm, or other direct, indirect, incidental, special, or consequential damages arising from the use of or inability to use the extension or OpenAI’s services.

17. OpenAI terms and policies

The site operator should regularly consult OpenAI’s official documents, as they may change over time:

• OpenAI Service Agreement
• OpenAI Usage Policies
• Professional Services and API Privacy
• API Data Controls and Retention

18. Legal notice

This document provides general information about ConvoViva’s operation and associated risks. It does not constitute legal advice.

Agencies, developers, businesses, and other organizations that deploy the extension for their clients should consult legal counsel to prepare terms of use, a privacy policy, and consent mechanisms tailored to their situation.