ResponsibleAIPolicy

This policy applies to all AI-assisted features offered through the BlueHive platform, including the assistant known as "Bea," as well as search, summarization, recommendation, and routing features that use machine learning. It applies to everyone who uses these features, including employers, providers, and examinees.

BlueHive operates AI features under the following commitments:

• No training on your data. BlueHive does not use Protected Health Information (PHI) or identifiable customer data to train AI or machine-learning models.
• Vetted providers, controlled retention. We use a small set of established AI subprocessors (such as OpenAI and Anthropic) and configure them with zero-retention or short-retention controls where available. Current AI subprocessors are disclosed on our subprocessor list.
• BAA before PHI. PHI is not transmitted to an AI provider unless an executed Business Associate Agreement is in place and the workflow is expressly designed to handle it.
• Human oversight. AI features assist people; they do not make medical, employment, or compliance decisions. A qualified person remains responsible for any regulated decision.
• Transparency. Our AI data handling is described here, in our Privacy Policy, and in our public subprocessor list.

BlueHive may make available chatbot, search, summarization, recommendation, and other AI features (including the assistant known as "Bea"). These features are intended to help you navigate the platform, find general information, and move work forward more quickly.

Bea is BlueHive’s in-platform assistant for employers, providers, and examinees. Ozwell (formerly BlueHive AI) is our separate clinical documentation assistant for providers, available at ozwell.ai and governed by its own terms.

When using AI features, you agree that you will:

• Not submit PHI, full Social Security numbers, payment-card numbers, or other highly sensitive information through public chat or AI tools unless BlueHive has expressly directed you to do so through a secure workflow.
• Independently verify any output before relying on it for a regulatory, clinical, employment, or business decision.
• Not use AI features to attempt to reverse-engineer, extract training data from, overload, or otherwise misuse the platform or its underlying models.
• Understand that BlueHive may log, monitor, audit, suspend, modify, or discontinue any AI feature at any time.

AI features may reflect gaps or errors in their underlying data and can produce results that appear confident but are wrong. You and your organization remain responsible for the decisions you make.

Our handling of personal information and PHI in connection with AI features is governed by our Privacy Policy. The third parties that may process data to power AI features are disclosed on our subprocessor list, which we update with advance notice as described there.

BlueHive may monitor and audit use of AI features to maintain security, quality, and compliance. We may update this policy as our features and the broader AI landscape evolve; material changes will be reflected by the "last modified" date above. Questions can be directed to our team through the contact page.