Artificial Intelligence (AI) Policy

Key Points

• General principles:
• Human oversight is required.
• Verify AI outputs for accuracy;
• Mitigate and monitor bias;
• Protect patient privacy/data security;
• Use AI responsibly/ethically;
• Report issues via the IT Service Desk.
• Access & approval:
• Only AI systems approved through Technology Governance, and the AI Governance Committee may be used; completion of mandatory education is required before access is granted.
• Restrictions:
• Unapproved AI systems are blocked on Harris Health devices and may not be used on personal devices for Harris Health work.
• Data Minimization & Compliance:
• Limit AI systems to necessary data/functions per Global Data Classification;
• Complete required regulatory reporting;
• Comply with applicable laws, Harris Health policies, and the forthcoming Texas AI Code of Ethics.
• Prohibited actions:
• Unauthorized access;
• Data manipulation;
• Privacy violations;
• Biased/discriminatory deployment;
• Unethical use;
• Relying on AI without human oversight;
• Security/privacy breaches;
• Developing/fine-tuning unapproved AI with Harris Health data/systems.