AI and Autonomous Operations Schedule

1. What GBOS AI does

GBOS provides AI-assisted recommendations and can run autonomous agents that take actions within your workspace. AI output is decision support, not professional advice; you remain responsible for reviewing it before relying on it. [SOLICITOR TO CONFIRM] the precise “not professional advice” positioning.

2. The authority boundary (implemented control)

Every intended agent action is classified before anything happens: a halted agent (kill switch) can take no action; a consequential action (spend, external effect, or sensitive-data action) is held for human approval by default; an action above the agent’s configured authority limit is held for a human; only in-limit, non-consequential actions auto-execute — and even those are audited with agent provenance.

You configure each agent’s authority limit and whether it may auto-execute consequential actions (off by default). You are responsible for the limits and approvals you set.

3. Data used by AI

AI features operate on your workspace data under your instruction. GBOS does not use Customer Data to train general-purpose models by default. [SOLICITOR TO CONFIRM] any exception and the consent/controls it would require. Third-party model providers, where used, are TBD-vendor and will be listed as sub-processors before activation.

4. Human oversight

High-risk and consequential actions require human approval; approvals are governed and audited. [SOLICITOR TO CONFIRM] the mandatory-approval action classes and warnings for financial, employment, legal, health/safety or otherwise regulated decisions.