# Quick-Reference Card: Responsible AI for All Staff

<p class="meta">FAST-17 · 7formations · Instructor: Wadood Chaudhary · Keep at your workstation · Not legal advice · Based in part on Svekis, AI Literacy for All Employees (2026), chapters 8 to 12 and 14</p>

## 1. Four principles

**Fairness · Accountability · Transparency · Privacy.** The tool never owns the outcome; people do.

## 2. Five-step ethical check

Stakeholders → Risks → Transparency → Oversight → Compliance

## 3. Bias types

Historical · Representation · Measurement · Evaluation. Watch for stand-ins (zip code, flags, gaps). **See a pattern? Write it down and report it.**

## 4. Data: OK, mask, never

| OK | Mask first | Never (in any AI tool) |
| --- | --- | --- |
| Public info, blank templates, own non-sensitive notes | Names, addresses, client names → [PLACEHOLDERS] | Name + SSN, license or state ID, account or card number (G.L. c. 93H, 201 CMR 17.00); health data; passwords; controlled technical data |

Approved tools and work accounts only. Check memory and training settings.

## 5. Scam check

Urgent + secret + money or data = **stop**. Verify by a known number or in person. Report it.

## 6. Policy in one line each

Approved tools: ____________ · Needs approval: hiring, discipline, pay, safety sign-off · Document: "AI-assisted; checked by ____" · Report to: ____________

## 7. My 90-day plan

Skill 1: ____________ · Skill 2: ____________ · Check-in: ____/____ · Reviewer: ____________