4th International Conference on
AI in Healthcare, Industry & Regulatory Affairs
May 25–27, 2027 | Toronto, Canada
Four Points by Sheraton Toronto Hotel
Address: Mississauga, 6090 Dixie Road, Mississauga, ON, L5T 1A6, Canada
AIHIRA 2027
ISO 14971 & AI Risk Management
ISO 14971 focuses on risk management for medical devices.
For AI, risk management becomes more complex because AI can behave differently depending on:
- Data
- Patient population
- Clinical environment
- Input quality
- Model limitations
- Updates
- Changes in real-world use
An AI risk-management process can therefore examine:
Hazard → AI Failure Mode → Probability → Severity → Risk Control → Verification → Residual Risk → Post-Market Monitoring
Examples of AI risks:
- Incorrect diagnosis
- False negative
- False positive
- Biased output
- Model drift
- Data corruption
- Unexpected model behavior
- Automation bias
- Human-AI interaction failures
- Cybersecurity-related AI risks