Report prepared by the Montreal AI Ethics Institute In Response to Mila's Proposal for a Contact Tracing App

Diversity and equity gaps: The report identifies concerns that the COVI app's design may not adequately consider diverse populations, suggesting that contact tracing solutions risk excluding or disadvantaging certain groups if accessibility and cultural factors aren't carefully addressed during development.

Unexamined assumptions about user behavior: The analysis questions whether the app's designers sufficiently tested their assumptions about how people will actually use the app and whether users will follow recommended actions based on risk information—a critical gap since real-world adoption and compliance determine whether contact tracing can be effective.

Transparency and accountability concerns: The report raises unanswered questions about how the app's risk calculations work, who is responsible if the app provides incorrect information, and how users can understand or challenge the system's decisions—issues essential for public trust in health technologies.

Security and data protection gaps: Despite claims that data would be safeguarded from governments and companies, the report identifies unresolved security questions, suggesting that privacy protections require more rigorous scrutiny and clearer safeguards before deployment.

Broader implications for AI governance: By analyzing COVI as a case study, the report demonstrates that responsible AI development in high-stakes domains like public health requires not just technical excellence but also careful attention to ethical considerations, equity, and accountability mechanisms that go beyond the original design team's analysis.