CARESHIP

AI Disclosure and Transparency Policy

Operated by Oneloop Inc. (a Delaware corporation), doing business as Careship

Effective Date: May 31, 2026

Oneloop Inc., doing business as Careship (“Oneloop,” “Careship,” “we,” or “us”), believes that people deserve to know when they are interacting with artificial intelligence. This AI Disclosure and Transparency Policy explains how our AI agent, Ela, works, how it is identified to the public, and the responsibilities that Customers accept when deploying it. This Policy is incorporated into our Terms and Conditions, End User License Agreement, and Acceptable Use Policy.

1. What Ela Is

Ela is an artificial intelligence agent that engages website visitors and callers on behalf of a Customer’s direct primary care practice. Ela uses machine learning models, including models provided by Anthropic, together with voice technology provided by ElevenLabs and telephony provided by Twilio, to hold conversations through chat, text, voice, and calls.

Ela is software. It is not a human, not a licensed healthcare professional, and not a provider of medical advice.

2. Disclosure to Members

Ela is designed to identify itself as an artificial intelligence agent during its interactions. We require that this disclosure remain enabled. Customers must not configure, instruct, or use Ela in a way that causes it to deny being AI or to present itself as a human.

Typical disclosure takes the form of an introduction such as identifying itself by name as an AI assistant for the named practice at the start of an interaction. Customers are responsible for ensuring the disclosure is accurate to their practice and remains active.

3. How Ela Handles Conversations

• Ela responds based on the knowledge, links, and instructions a Customer configures, along with general model capabilities.
• Conversations, including voice calls, may be recorded, transcribed, summarized, and analyzed for sentiment to provide the service to the Customer.
• Ela is designed to hand off to a human when a situation falls outside its scope or when a Member requests it.
• Ela does not make clinical decisions and is not a diagnostic tool.

4. Accuracy and Limitations

Artificial intelligence can produce output that is inaccurate, incomplete, outdated, or unsuitable for a particular situation. While we design Ela to be helpful and to reduce errors, we do not guarantee that its output will be correct or appropriate in every case.

Members should not rely on Ela for medical advice, diagnosis, treatment, or emergency assistance. In an emergency, Members should contact emergency services or a qualified healthcare professional directly.

5. Customer Responsibilities

Customers who deploy Ela agree to:

• Keep Ela’s AI disclosure enabled and accurate at all times.
• Refrain from configuring Ela to impersonate a human or to mislead Members.
• Refrain from using Ela to provide medical advice, diagnosis, or treatment.
• Obtain all consents required for recording, calling, and messaging Members.
• Provide accurate practice information so Ela does not make false or unsubstantiated claims.
• Supervise Ela’s operation and review its output before relying on it.

These responsibilities support compliance with consumer protection expectations, including those of the Federal Trade Commission regarding the disclosure of AI and the avoidance of deceptive practices. Customers, not Oneloop, are responsible for how they deploy and configure Ela.

6. Data and Privacy

Information collected through Ela is handled in accordance with our Privacy Policy. We do not use Member conversations to train our own foundation models, and we do not permit our AI Sub-processor to use Customer or Member data to train its general models, consistent with that Sub-processor’s commercial terms. Our Sub-processors are disclosed in the Privacy Policy.

7. Human Oversight and Escalation

Ela is intended to assist, not to replace, human judgment. Customers remain responsible for the care, communications, and decisions of their practice. Ela is designed to escalate to a human where appropriate, and Customers should maintain a path for Members to reach a human.

8. Changes to This Policy

As AI technology and the law evolve, we may update this Policy. Material changes will be posted with an updated effective date. Continued use of the Platform after changes take effect constitutes acceptance.

9. Contact

Questions about how Ela works or about this Policy may be directed to:

Oneloop Inc., doing business as Careship

Email: hello@careshiphealth.com

Phone: +1 323 310 1076