Acceptable Use Policy

Introduction

Welcome to Sorcero. This Acceptable Use Policy ("Policy") outlines certain obligations of Users with respect to their use of the Sorcero AI Platform ("Platform"). By accessing or using the Platform, Users agree to comply with this Policy. This Policy is designed to ensure the responsible and ethical use of the Platform, protect data privacy, and maintain compliance with applicable healthcare regulations.

Platform Usage Guidelines

Please note that AI-generated results are intended to support, not replace, your professional expertise and judgment. Always use your professional judgment in evaluating any Platform outputs. Please be sure to review and comply with our Acceptable Use Policy with respect to all Platform queries. Without limiting the generality of the foregoing, do not submit prompts that would not be appropriate for a Medical function, for example:

Requesting recommendations to support commercial or promotional objectives.
Requesting information or recommendations regarding transfers of value to HCPs.
Requesting information or advice on how to treat an individual patient.
Providing any personally identifiable patient information in Platform queries.

Understanding Sorcero's Approach to AI

Sorcero is committed to ensuring meaningful human oversight over the design, development, deployment, and operation of our Platform. Our primary goal is to mitigate risks associated with AI, ensure ethical considerations are addressed, and maintain accountability.
Sorcero's approach to human oversight is guided by principles such as:

Accountability: Humans are ultimately accountable for the decisions and actions of AI systems.
Transparency: The capabilities and limitations of AI systems, as well as their decision-making processes, should be understandable.
Fairness and Non-Discrimination: AI systems should be designed and operated to avoid unfair bias and discriminatory outcomes.
Safety and Reliability: AI systems must be safe, reliable, and perform as intended. Human oversight is critical to identifying and mitigating potential safety risks.
Privacy: AI systems must respect and protect individual privacy.
Security: AI systems and the data they process must be protected from unauthorized access, use, disclosure, alteration, or destruction.
Human Control: Appropriate levels of human control and intervention must be maintained, especially for high-risk applications.

Your Responsibilities

As a Platform User, you are an essential part of our human oversight framework. You are responsible for understanding the system's purpose, limitations, and how to escalate issues or override automated decisions when necessary.

Human Oversight is Essential: All AI-generated results are intended to support, not replace, your professional expertise and judgment. You are always accountable for any insights or recommendations you choose to act upon in a compliant and ethical manner consistent with all Gilead principles and policies.
No Commercial or Promotional Recommendations: The Platform's primary purpose and function is to assist the user in extracting critical insights from medical data, it is not intended to serve a commercial function. Do not create outputs designed to drive sales or support Commercial objectives.
Avoid Improper Transfer of Value (Quid Pro Quo): Recommendations for HCP activities (speaker, advisor, etc.) must be based only on skills, interests, and experience, not as an inducement to prescribe. Do not provide or solicit information regarding HCP compensation.
Do Not Practice Medicine: The AI is explicitly trained to refuse to provide medical advice, diagnosis, or treatment recommendations for any individual, real or hypothetical. Do not use the system to give advice on how to treat individual patients. Aggregated, population-level insights are permissible.
Protect Patient Information (PHI/PII): Users must ensure that any data input into the Platform is anonymized to the extent possible to protect patient privacy. Personal identifiers should be removed unless absolutely necessary for the intended use. Do not include any personally identifiable patient information in responses. Our system is designed to prohibit PII from being ingested. Aggregated/anonymized data and HCP information from profiles are permissible.
Review and Validate: Users are responsible for ensuring the accuracy and completeness of any information input into the Platform. In addition, always critically review and validate AI-generated outputs for accuracy, compliance, and ethical considerations before use. If there is any uncertainty about compliance, assume it is non-compliant until verified.

Sorcero's Commitment to Fairness and Bias Mitigation

At Sorcero, we are dedicated to developing and deploying AI systems that are as fair, equitable, and unbiased as possible. We recognize the critical importance of this in the life sciences sector. Our goal is to ensure our AI solutions are trustworthy, transparent, and serve all users responsibly.
Our strategy for promoting fairness and minimizing bias includes:

Rigorous Data Management: Employing systematic data cleaning and transformation and actively striving to incorporate a wide range of sources, perspectives, and contexts within our training datasets.
Proactive Bias Detection and Mitigation: Implementing techniques to detect and mitigate bias at various stages of the AI lifecycle.
Transparency and Interpretability: Prioritizing Explainable AI (XAI) to help users understand how models arrive at outputs, and maintaining clear documentation.
Ethical Framework and Human Oversight: Guiding development by ethical principles emphasizing inclusivity, fairness, and accountability, and integrating human judgment to review and validate AI outputs.
Continuous Validation and Improvement: Regularly evaluating AI models against benchmarks and incorporating user feedback.