Control AI features & execution modes
The Settings tab lets you switch individual AI features on or off and control which execution mode each runs in. Kai Hub is admin-only. Settings has five inner sub-tabs: 🧩 AI Features · ⚡ Modes · 🛡️ Employee Reserve · 📏 Caps & Limits · 🔔 Alerts. This article covers the first two.
AI Features — per-module toggles
Open Settings → AI Features to see AI features · per-module toggles, with a Save button (toast: Feature settings saved). This is the enable/disable-AI-per-feature control.
Features are grouped by module — Kai Copilot ✨, Workforce Intelligence 📊, ATS · Hiring 🧲, Letters ✉️, Payroll 💸, and Performance 🎯. Each feature shows its mode badge, {price} cr per action, and an on/off toggle (toast: {feature} enabled or {feature} disabled).
Modes — execution modes
Open Settings → Modes to see Execution modes. Three modes are fixed to Genius or Swift, and one is toggleable:
- Kai Copilot · Genius responses ◆ Genius — "Frontier model + extended tool use for admin copilot · 7 cr/msg (token-capped)" — toggleable.
- Workforce Intelligence ◆ Genius — locked on — "Always Genius — advanced analytics is the promise" (toast: WI always runs on Kai Genius).
- Payroll Auto-Pilot ◆ Genius — locked on — "Always Genius — high-stakes reasoning" (toast: Payroll Auto-Pilot always runs on Kai Genius).
- Everything else ⚡ Swift — locked — "Fast, low-cost models · employee flows are always Swift" (toast: Swift is the platform default).
These are the three features reserved for Kai Genius — Workforce Intelligence and Payroll Auto-Pilot are locked on, while Kai Copilot's Genius responses can be toggled. Everything else runs on Kai Swift, and employee flows are always Swift.
For the full list of what each feature costs, see Understand credits, modes & usage metering.
When to contact support
If a feature toggle or a locked mode does not behave as described here, contact support with the module and feature name.