About TIERRA
Committed to Ethical AI Innovation
TIERRA is a collaborative space that seeks to harness the knowledge and power of stakeholders of health AI. Using the diverse expertise of academic faculty and students, industry professionals, policy and decision-makers, and community members, TIERRA aims to create collaborations, research, roadmaps, and resources to inform best practices and policies on the development and application of AI in healthcare settings.
What are artificial intelligence (AI) tools?
AI tools are created by training computer programs to recognize patterns in large amounts of data like images, speech or text. The programs are given many examples, tested on how well they perform, and then given feedback so that they can improve their generative and predictive or diagnostic capabilities.
AI Use Cases
| Predictive AI | Generative AI | |
| Personal | Track medications, send health alerts | Generate personalized medication reminders |
| Healthcare Providers | Predict health needs and develop personalized treatment plans | Draft email communication to patients |
| Insurance Companies | Predict healthcare needs and future costs | Automate claims processing |
| Public/Population Health | Forecast spread of disease | Produce health trend reports |
AI Risks in Healthcare
- Trust and Transparency: Patients may not know that AI is being used.
- Equity: AI tools may not work for segments of a population or less data-represented groups, which can lead to bias or errors.
- Reliance on AI for Clinical Decisions: AI may help doctors do their jobs more effectively. However, doctors and clinics using AI may become too dependent on the technology, leading to blind spots in situations needing human judgement.
- Access and Who Benefits: Using high quality AI technology in healthcare can be expensive. Not all people or healthcare organizations will have access to these tools or may opt for less expensive, but inferior AI technology creating an “AI Digital Divide”.
- Oversight for Safe, Effective, and Ethical Use: As AI becomes more integrated in healthcare, there is a need for appropriate oversight to make sure that AI is used safely, effectively, and ethically. Robust governing principles and regulatory frameworks around AI are lacking at the national and state levels.
- Privacy and Security: AI needs an immense amount of data to work accurately and effectively. This data often comes from and goes to public agencies, private companies or other organizations outside of the healthcare system. While security measures are taken to safeguard the data, there remains a risk that data security is compromised (i.e., hacked or systems breakdown)
What are trustworthy health and AI systems?
Trustworthiness, as described by the Association of American Medical Colleges’ Center for Health Justice, is a “necessary precursor to trust that is rooted in honesty and honors lived experience.” It is also key to a successful patient-provider partnership. Trustworthiness is fundamental, human-centered, and earned based on actions taken to address unmet needs.
