AI Policy
At BME Strategies (BME), our mission is centered on driving impactful, community-focused solutions. As artificial intelligence (AI) technologies evolve, BME supports and encourages their responsible integration to enhance operational efficiency, streamline workflows, and elevate service delivery.
This public AI Policy outlines our commitment to ethical, secure, and equitable AI deployment. These guidelines were thoughtfully developed by BME personnel through a series of dedicated committee discussions and collaborative meetings, ensuring our approach was shaped by diverse perspectives across the organization. It establishes the principles and operational boundaries that govern how BME utilizes AI across all projects, client engagements, and internal workflows.
Our Core Principles: The C.A.R.E. Framework
To ensure we harness AI’s potential while maintaining BME’s standards of excellence, our AI council has established four core values:
- C – Centering You: AI is used to amplify human creativity, insight, and strategic thinking—never to replace them. Lived experiences and human expertise remain at the heart of our deliverables.
- A – Accountability: BME retains complete responsibility for all work products. AI outputs are treated as raw material and must undergo rigorous human review, fact-checking, and editing before finalization. Unverified or raw AI generation is strictly prohibited in client deliverables.
- R – Respect Privacy: Protecting client, partner, and organizational data is paramount. Proprietary, sensitive, or protected information is strictly safeguarded and processed only within secure, vetted environments.
- E – Equity, Always: AI outputs are systematically evaluated through an equity lens to identify and dismantle algorithmic bias, ensure accessible language, and maintain alignment with racial equity principles and Culturally and Linguistically Appropriate Services (CLAS) standards.
Data Privacy, Security & Tool Governance
Accepted & Secure AI Environments
- Approved Enterprise Tools: To safeguard client information, BME exclusively utilizes AI tools that are SOC II Type II compliant. These third-party audited systems ensure that data is not retained to train AI models. Under these operational environments, input data is protected and strictly prohibited from being used to train public or third-party underlying models.
- Client-Specific Rules: BME strictly adheres to all client-specific AI constraints and policies (such as prohibitions against AI note-taking applications during confidential meetings).
Prohibited Uses & Data Protection Rules
- No Unvetted Tools: Personnel are prohibited from using unverified, third-party AI tools or platforms on devices that access or host BME or client data.
- Data Restrictions: Sensitive information—including Protected Health Information (PHI), personally identifiable information (PII), confidential client data, and BME proprietary frameworks—must never be uploaded into public or non-secure commercial AI models.
- No Blind Attribution or Plagiarism: AI-assisted research and content must maintain strict copyright integrity. Original work by human authors or creators must receive proper attribution.
Transparency & Ethical Oversight
BME actively audits AI-assisted workflows to guard against bias, stereotyping, or inequitable outcomes. When leveraging AI for analysis or content generation, team members are required to:
- Validate all source data and independently verify factual accuracy.
- Evaluate whether recommendations, outputs, or language disproportionately affect groups based on race, ethnicity, gender, ability, language, or socioeconomic status.
- Ensure community voices and diverse lived experiences are respected and accurately represented.
Governance & Policy Maintenance
AI technology and regulatory standards change rapidly. BME maintains internal governance to monitor compliance, assess software, and update organizational safeguards.
- Review Cadence: This policy is subject to review as technology or workflows are evaluated or immediately following significant regulatory or technical shifts.
- Incident Response: Any suspected data security or privacy issue involving AI must be immediately escalated in accordance with BME’s formal Data Incident Protocols.