AI Usage Guidelines at Brainforge

This document defines how AI should and should not be used at Brainforge.
AI is a productivity tool, not a decision-maker.


1. Approved Uses of AI

AI may be used to:

  • Draft documents
  • Summarize transcripts
  • Rewrite for clarity and tone
  • Organize notes
  • Generate first-pass structures
  • Identify gaps or inconsistencies

AI accelerates thinking but does not replace judgment.


2. Disallowed Uses of AI

AI must NOT be used to:

  • Invent facts or requirements
  • Make legal or compliance decisions
  • Finalize pricing without review
  • Commit code or docs without human review
  • Replace SME input

3. Human Accountability

Every AI-generated output must be:

  • reviewed by a human
  • fact-checked
  • aligned with real client discussions

The human is always responsible for correctness.


4. Avoid Hallucinations

AI may confidently generate incorrect information.

To reduce risk:

  • Provide clear context
  • Use templates
  • Run accuracy checks
  • Ask AI to flag uncertainty
  • Compare against source materials

5. Sensitive Information

Do not:

  • paste confidential data unnecessarily
  • expose private client information outside approved repos
  • share sensitive outputs without review

Always follow internal data handling policies.


6. Iterative Improvement

Use AI iteratively:

  • generate
  • review
  • refine
  • review again

Quality improves through iteration.


7. When to Escalate

Escalate to a PM or lead if:

  • AI outputs conflict with client expectations
  • requirements are unclear
  • scope boundaries are uncertain
  • sensitive decisions are involved

8. AI Is a Tool, Not an Authority

AI helps you work faster.
It does not replace:

  • judgment
  • accountability
  • responsibility
  • expertise

Use AI thoughtfully and deliberately.