Your team's training data belongs to your organization.

The people in your organization (name, email, job and roles), the training assigned to them, their test and scenario answers and results, the skill evidence built from those results, supervisors' review decisions and feedback, and the training content your organization uses or creates.

Important changes, such as approving content, grading an answer or changing someone's role, are recorded in an audit log that admins can read.

Everything works without AI. If you turn it on, an admin adds your organization's own OpenAI API key (OpenAI is the supported provider), chooses which AI features are on and sets a monthly spending limit. gr0wpath estimates spending from each request and pauses AI for the rest of the month when the estimate reaches the limit.

When an AI feature runs, the text that task needs is sent to that provider. For example, a written answer is sent so the AI can suggest a grade.

AI-drafted lessons, questions, scenarios and courses stay drafts until a person approves them. A person can approve one item at a time or several at once, and the record shows a bulk approval as a bulk approval rather than an item-by-item review. AI-generated pictures are only allowed as decorative illustrations; an admin can choose to approve those automatically, and that is recorded too.

By default, AI-suggested grades on tests and exams don't count until a supervisor reviews and approves them. An admin can change that setting. In practice mode, learners usually see the AI's feedback straight away and it updates their practice progress, but AI-graded answers don't count toward verified skill ratings until a person approves them. Answers the AI isn't confident about always go to a person.

gr0wpath doesn't decide who is hired, promoted, disciplined or let go. Results, readiness checks and hiring assessments are evidence for a person to weigh, alongside everything else they know. Written and practical answers in a hiring test are scored by a person.

Results describe recent work, never people. Skills that haven't been tested show as "Not tested yet", never as a zero.

Employees see their own results and the evidence behind each skill rating. If they think a rating is wrong, they can ask for a correction or a fresh assessment. A supervisor responds, and the employee sees the response.

Completion records and certificates are internal records of your organization's training, not accreditation.

Employees see their own training and results. Managers see the results of the teams they manage. Admins manage settings, content and people. You can sign in with Microsoft Entra ID single sign-on, which follows your own Entra multi-factor policies, and require two-step verification for chosen roles when people sign in with a password.