AI Transparency Statement

This AI Transparency Statement is published in compliance with California Senate Bill 942 (Artificial Intelligence Transparency Act), effective January 1, 2026, and as a general commitment to transparency in AI-powered interactions.

1. What Is OmniBotX?

OmniBotX is a B2B SaaS multilingual AI chatbot platform that enables businesses to deploy conversational AI assistants on their websites. The platform is developed and operated by Atronus Technologies, Inc.

2. AI Technology Used

OmniBotX is powered by the following AI technologies:

• Anthropic Claude (claude-sonnet-4-6 / claude-opus-4-6) — Large language model used for generating conversational responses. Developed by Anthropic, PBC (San Francisco, CA).
• Google Cloud Translation API — Used for language detection and cross-language translation of conversations. Developed by Google LLC.

Atronus Technologies does not train proprietary AI models. The underlying AI capabilities are provided by the above subprocessors under their respective terms of service.

3. CA SB 942 Compliance

In compliance with California SB 942, OmniBotX implements the following disclosures:

• AI Identity Disclosure: Every OmniBotX bot will always identify itself as an AI assistant when sincerely asked by a user. Bots are instructed never to claim to be human.
• Bot Name Transparency: Each deployed bot has a visible name displayed in the chat header. The bot name is configured by the Subscriber.
• Powered-By Attribution: Where displayed, the "Powered by OmniBotX" attribution identifies the AI platform provider.
• No Deceptive Impersonation: OmniBotX is configured to never impersonate a specific individual, government entity, or claim capabilities it does not have.

4. How AI Responses Are Generated

When a user sends a message to an OmniBotX-powered bot:

1. The message is sent to Atronus's secure Cloudflare Worker
2. The worker applies content moderation filters to screen for harmful content
3. The message, along with the Subscriber's business context (name, services, FAQs), is sent to Anthropic's Claude API
4. Claude generates a response based on the business context and conversation history
5. The response is returned to the user in their detected language

OmniBotX does not use AI to make automated decisions about individuals that produce legal or similarly significant effects.

5. AI Limitations

Users should be aware of the following AI limitations:

• AI responses may occasionally be inaccurate, incomplete, or outdated
• AI responses do not constitute legal, medical, financial, or professional advice
• The AI may not understand highly technical, domain-specific, or ambiguous queries
• Response quality depends on the accuracy of the business information provided by the Subscriber

6. Content Moderation

OmniBotX employs automated content moderation that screens incoming messages for profanity, threats, sexual content, spam, and prompt injection attempts before they reach the AI model. Moderated messages receive a polite deflection response and are logged for security purposes.

7. Subscriber Obligations

Businesses deploying OmniBotX ("Subscribers") are required under our End User License Agreement to:

• Inform End Users that they are interacting with an AI system
• Not configure the bot to impersonate a specific human individual
• Comply with CA SB 942 and all applicable AI transparency laws in their jurisdiction
• Not use OmniBotX to deceive users about the nature of the interaction

8. User Rights Regarding AI

You have the right to:

• Ask the OmniBotX bot at any time whether you are speaking with an AI — it will always confirm it is an AI
• Request that your conversation data be deleted by contacting the Subscriber or Atronus directly
• Opt out of AI interaction by closing the chat widget

9. Colorado Artificial Intelligence Act (CAIA) Compliance

The Colorado Artificial Intelligence Act (Colorado Revised Statutes § 6-1-1701 et seq.) takes effect June 30, 2026 and imposes transparency and non-discrimination obligations on deployers of "high-risk" AI systems.

9.1 Is OmniBotX a High-Risk AI System?

Under the CAIA, a "high-risk AI system" is one that makes or substantially influences "consequential decisions" affecting consumers in areas such as employment, education, financial services, healthcare, housing, or legal services. OmniBotX is a general-purpose business chatbot designed to answer customer questions and capture leads. It is not designed to make consequential decisions as defined by the CAIA.

However, Subscribers who deploy OmniBotX in contexts that may involve consequential decisions (for example, financial services qualification, insurance, or housing inquiries) should independently assess whether the CAIA's high-risk AI obligations apply to their specific deployment.

9.2 Transparency Disclosures (All AI Systems)

Regardless of risk classification, the CAIA requires that AI systems intended to interact with consumers disclose their AI nature unless it is obvious to a reasonable person. OmniBotX complies with this requirement as follows:

• Every OmniBotX bot identifies itself as an AI assistant when sincerely asked
• Bot names are displayed visibly in the chat header at all times
• The OmniBotX platform does not simulate human identity or claim human capabilities

9.3 Algorithmic Discrimination

Atronus Technologies has designed OmniBotX to avoid algorithmic discrimination. The AI responds based on business context provided by the Subscriber and the content of the user's message. OmniBotX does not use protected characteristics (race, gender, religion, national origin, disability, etc.) as factors in generating responses.

9.4 Subscriber Obligations Under CAIA

Subscribers deploying OmniBotX in Colorado or to Colorado consumers must:

• Assess whether their specific use case constitutes a "high-risk" deployment under CAIA § 6-1-1702
• If high-risk: provide the required consumer notices, maintain risk management policies, and submit annual impact assessments to the Colorado Attorney General as required
• Not configure OmniBotX to make or substantially influence consequential decisions without appropriate human oversight and consumer disclosure

10. Governing Law

This disclosure is made pursuant to California SB 942 (Chapter 762, Statutes of 2024), effective January 1, 2026. It also addresses the EU AI Act's transparency requirements for AI systems that interact with natural persons (Article 52).