What Happens If the AI Provider Goes Down or Changes?

Changes to plan for

• —Outages
• —Model changes
• —API changes
• —Pricing changes
• —Usage limits
• —Feature changes
• —Provider policy changes

Design approaches

• —Graceful failure — the rest of the business keeps working
• —Retry logic for temporary errors
• —Queues, so work waits instead of being lost
• —Monitoring and alerts
• —Human fallback for important work
• —Provider abstraction where justified — isolating the AI provider so it can be changed
• —Alternative model or provider options where appropriate
• —Keeping company data in company systems, separate from the AI model itself

Switching is not always instant

Not every system can switch providers instantly, and different models can behave differently. How much resilience to build depends on how critical the workflow is.