AI Use & Transparency Policy

How Pluse uses artificial intelligence in the Lux assistant.

Last updated: May 15, 2026  ·  Version 1.0

1. Summary

Pluse uses artificial intelligence to power Lux, an in-product assistant designed to help small businesses understand their data and complete common tasks faster. This policy explains, in plain English, what Lux is, how it works, what data it sees, what it does not do, and how you can opt out.

The short version. Lux is powered by the Claude Sonnet 4.5 language model (made by Anthropic), accessed through DigitalOcean’s Generative AI Platform. Your conversations with Lux are processed by DigitalOcean on our behalf to generate responses. Your data is not used to train any general-purpose AI model. Lux can be disabled. AI outputs are not legal, tax, or financial advice.

2. What Lux Is

Lux is an AI-powered assistant integrated into the Pluse platform. It can answer questions about your business data, summarize trends, draft invoice descriptions, suggest categorizations, and respond to natural-language questions about the Service. Lux is identified throughout the product by the sunburst mark and the name “Lux.”

Lux is provided as a convenience tool, not a substitute for professional advice. See Section 6 (Limitations) below.

3. AI Provider & Model

Lux is built on Claude Sonnet 4.5, a large language model created by Anthropic, PBC (a Delaware public benefit corporation headquartered in San Francisco, California). We do not access Claude directly from Anthropic. Instead, Lux interacts with the model through the DigitalOcean Generative AI Platform, a managed AI inference service offered by DigitalOcean, LLC on which the Claude model family is deployed.

The data-processing relationship is therefore structured as follows:

The specific Claude model version may change over time as DigitalOcean’s GenAI Platform offers new releases. We choose model versions based on quality, latency, cost, and safety characteristics. The current production model is Claude Sonnet 4.5.

4. What Data Lux Processes

When you interact with Lux, the following data is sent to the DigitalOcean Generative AI Platform on our behalf to generate a response:

The following data is not sent to the AI provider unless you include it explicitly in a prompt:

5. Training Data & Model Improvement

This is the question we hear most often:

Your prompts and Lux’s responses are not used to train any general-purpose AI model. Under the DigitalOcean Generative AI Platform terms of service, customer inputs and model outputs processed through the platform are not used to train, retrain, or fine-tune any foundation model on customer data. This commitment runs directly from Pluse to DigitalOcean. Anthropic, the underlying model creator, is not a party to that commitment, but Anthropic’s contractual relationship with DigitalOcean for the operation of the GenAI Platform governs whether and how Anthropic accesses inference traffic on that platform.

Pluse may aggregate de-identified usage metrics — for example, the count of Lux conversations per day, average response length, or common task categories — to improve our system prompts, refine product behavior, and measure usefulness. To be clear, these metrics are used to tune how Pluse builds Lux, not to train or fine-tune any AI model. The metrics do not contain the content of any conversation and are not personally identifiable.

If we ever change this policy — for instance, by introducing an opt-in feature where customer data is used to fine-tune a Pluse-specific model — we will (a) update this page, (b) require explicit opt-in, and (c) provide an opt-out at any time. We will never silently start training on customer data.

6. Limitations & Risks

AI outputs can be wrong. Large language models can produce confident, plausible-sounding answers that are factually incorrect (“hallucinations”), out of date, or inconsistent with your actual data. Always verify any AI-generated output before relying on it for an important business decision.

Specifically, Lux:

You should treat Lux as a knowledgeable but fallible assistant — useful as a starting point, not as the final word.

7. Human Oversight

Lux does not take consequential automated actions on your behalf without your explicit confirmation. Actions that change data — creating an invoice, recording a payment, sending an email, syncing to QuickBooks — require you to review and confirm before they execute. Lux is a copilot, not an autopilot.

Pluse engineering and trust-and-safety personnel may review a small sample of Lux conversations for safety, abuse prevention, and quality assurance. Such review is governed by our Access Controls Policy and applies the same access restrictions as any other customer data.

8. Your Consent & Opt-Out

You will be asked to consent to AI features at first use or, on iOS, during in-app onboarding. By using Lux you agree to the processing described in this policy and in our Privacy Policy.

You can disable Lux at any time:

Disabling Lux stops new conversations from being sent to the AI provider. Past conversation history stored in your account is retained per our Data Retention Policy and can be deleted by request.

9. Children & Sensitive Data

Lux is not intended for use by children under 13. Do not enter the personal information of children, health information, government identifiers (SSNs, passport numbers), or other sensitive data into Lux. If you do, we cannot guarantee it will be redacted before being transmitted to the AI provider.

10. Security

Conversations with Lux are transmitted over TLS 1.2 or higher between your device, the Pluse application servers, and the DigitalOcean Generative AI Platform. We apply the same technical and organizational security measures described in our Information Security Policy to AI-related data.

11. Regulatory Disclosures

11.1 FTC Guidance on AI

This policy is provided in alignment with US Federal Trade Commission guidance on AI transparency. Pluse does not misrepresent Lux’s capabilities, does not use Lux to engage in unfair or deceptive practices, and does not use Lux to make material decisions about you without disclosure.

11.2 California AI Transparency Act (SB-942)

The California AI Transparency Act applies to AI providers above a defined size threshold. Pluse is currently below that threshold, but this policy reflects our commitment to clear AI disclosure regardless of whether the Act applies. Content materially generated by Lux is identifiable as AI-generated within the product interface.

11.3 EU AI Act

Lux is a general-purpose conversational assistant and is classified as a limited-risk AI system under the EU AI Act. Pluse complies with applicable transparency obligations: users are informed they are interacting with an AI, and AI-generated content is identifiable as such.

12. Updates to This Policy

We will update this policy when our AI provider, model behavior, data-handling practices, or applicable law changes. Material changes will be reflected in the “Last updated” date and, for significant changes that affect your rights, communicated by in-product notice or email.

Questions about how Lux works?

For AI-specific questions or to report an AI output that is harmful, biased, or wrong, contact us.

Contact@pluse.to

This policy describes how we use AI in good faith and is informational, not contractual. Where this policy conflicts with our Terms of Service or Privacy Policy, those documents govern.

Emmber, Inc. is a Delaware corporation qualified to do business in the State of Florida. The “Pluse” name and the Pluse platform are operated by Emmber, Inc.

© 2024–2026 Emmber, Inc. All rights reserved. Pluse™ and Lux™ are trademarks of Emmber, Inc.