Human Oversight and Control
At Bigfoot Voice, AI operates under strict human oversight. Our AI assists — it never acts autonomously on matters that affect your service.
Human-in-the-Loop
- No Autonomous Actions — the AI cannot modify accounts, process payments, change service configurations, or make contractual commitments
- Human Review of All Requests — every support request opened through the AI creates a case that a human agent reviews and acts upon
- Human-Authored Content — all Knowledge Base content that drives AI behavior is written, reviewed, and approved by our team
- Curated Training — intent utterances and KB taxonomy are manually maintained, not auto-generated
Override and Control
- Instant Correction — problematic AI responses can be traced to their source KB article and corrected immediately
- System Controls — the AI can be disabled, retrained, or rolled back within minutes if issues are identified
- Your Choice — you can request human-only service at any time, for any reason
Review Schedule
AI accuracy is evaluated after every model update. Knowledge Base gaps are reviewed weekly. This governance policy is reviewed quarterly.