MAIN AI Playbook

Student AI Playbook

A practical guide to understand course rules, use AI to strengthen learning, protect personal information, check every important output, and submit work you can explain and defend.

Built on the Statewide AI Framework · See the alignment

Explore other AI playbooks

Strengthen the learning. Own the work.

Audience
Students using AI for coursework, study, and career preparation.
Produces
A checked course rule, verified work, disclosure where required, and a before-submission review.
Boundary
AI supports learning; students remain responsible for the work they submit.

Built on Mississippi’s Statewide AI Framework

The Mississippi Statewide AI Framework (August 2026) is the foundation for this playbook. It supplies shared priorities and a learning progression; this playbook applies them to student learning, coursework and career preparation. Both are voluntary. Applicable law, adopted standards and institutional authority remain controlling.

Read the August 2026 Statewide AI Framework (PDF)

Learning progression. High school and postsecondary entry points, adapted to the actual assignment. These activities support progression; completing a checklist does not establish a level.

Five priorities in practice

Use an approved, accessible learning route (1); verify and disclose assistance (2); protect personal and course information (3); practice skills you can explain and transfer (4); judge learning from evidence rather than use counts (5).

Priority numbers refer to literacy and access (1), ethics and critical thinking (2), privacy and governance (3), workforce readiness (4), and strategic alignment and measurable outcomes (5).

Selected Framework domains, practice and evidence
Framework domain Practice and evidence to retain
1. AI Foundations and Conceptual Understanding Explain why a fluent answer may be wrong; identify a limitation before relying on it.
2. Data Literacy and Data Stewardship Use only permitted information; record the assignment and account boundary.
3. Critical Evaluation of AI Outputs Open sources, recalculate and test; retain an error, correction and supporting evidence.
8. Human Agency, Judgment, and Restraint Do the required thinking and defend the final work; retain your reasoning and disclosure.

Keep the shared human-review boundary

Apply the Framework's human-review triggers (p. 35) before action: assessment or credentials; financial allocation or compensation; health or physical safety; discipline or employment; legal or compliance content; and sensitive or protected information. Its additional agent triggers cover autonomous system-of-record changes; persistent memory; tool access, credentials or external transactions; handoffs in safety-critical, regulated or student-facing processes; missing stop conditions or logged escalation to a named human owner; and institutional responses to synthetic media about identifiable people. Product or pilot approval does not replace action-specific review. Review the system and access before enabling them; review each proposed system-of-record change before execution. Advance permission cannot waive these triggers.

Use evidence within its limits

Use the crosswalk to select practice and retain observable evidence. It is MAIN's implementation mapping, not a validated proficiency test, credential or certification. Local rubrics and pilot measures alone do not establish statewide proficiency or causal impact. Recheck alignment when the Framework changes; record the edition, local owner and next review.

Source: August 2026 Framework, pp. 5-7 (priorities), 14-30 (domains and progression), and 35-37 (review, measurement and maintenance).

The short answer

How should students use AI?

Use AI as a learning aid, not a substitute for learning. Check the course rule before you begin, protect private information, ask the tool to help you practice or see alternatives, verify its claims, disclose or cite use when required, and submit only work you understand and can explain.

Your decision path: Name the task and skill you need to practice; check the rule, tool and information; try only permitted help; verify against real sources or your own work; keep, revise, or discard the result and disclose use as required; then ask whether it actually helped you learn. If the rule is unclear or the output cannot be checked, pause and ask your instructor. Discard output that invents evidence, exposes private information, or does required thinking for you.

Your instructor and institution set the rules for your course. This playbook is voluntary educational guidance; it does not replace a syllabus, assignment instructions, academic-integrity policy, accessibility process, professional standard, or law.

Read before you prompt

Know the rule for this activity

Do not assume one rule covers every course or every step. If instructions conflict or are unclear, pause and ask your instructor.

AI required

Use the named tool or approved alternative and demonstrate the skill the assignment is designed to teach.

AI permitted

You may use AI for stated purposes. Keep your own reasoning central and follow disclosure, citation, and verification rules.

AI limited

AI is allowed for some steps but not others. Record the boundary before you start and preserve evidence of your process.

AI prohibited

Complete the activity without AI. The purpose may be to measure your independent knowledge, writing, calculation, performance, or judgment.

Ask a precise question: “May I use [tool] for [specific step]? If yes, what should I disclose or cite, and what evidence of my own process should I submit?” Save the answer with the assignment instructions.

Practice help, course actions and applications

An approved tutor asking practice questions can support your learning. An agent logging into a course, answering a discussion or submitting an assessment as you raises a separate permission question. Before connecting an account, check what the tool can read, change or submit and whether each action is allowed. Do not let it replace performance you must demonstrate yourself. Preserve approved accessibility supports and ask your instructor when the boundary is unclear.

Check new rules before reusing AI-assisted work in an application, portfolio, interview or credential assessment. Permission for coursework does not establish permission for another recipient. Cornell’s graduate-admissions rules illustrate this difference; they are not general rules for MAIN learners.

Stanford academic AI guidance · Cornell graduate-admissions AI guidance

Practice more; outsource less

Six responsible student AI workflows

These patterns are starting points only when your course allows them. Give the minimum context needed, avoid private information, and compare outputs with course materials.

For a complete research-brief example, see the Student Playbook PDF, pages 6-7. It follows permission, search planning, independent source research, verification, writing and disclosure. Use the actual assignment rule and ask the instructor whenever a required condition remains unclear.

1. Explain and question

Ask for two explanations at different levels, then answer follow-up questions without help.

Try: “Explain [concept] using an example, then ask me three questions one at a time. Do not give the answer until I respond.”

2. Practice and retrieve

Turn your own notes into a practice set. Attempt it closed-book, then review mistakes against the original material.

Try: “Using only these notes, create five practice questions. Label anything the notes do not support.”

3. Brainstorm and revise

Generate alternatives, choose with reasons, write your own draft, and request focused feedback rather than a replacement answer.

Try: “Ask me questions that help narrow my claim. Do not write the paper or invent sources.”

4. Research support

Use AI to identify vocabulary, counterarguments, or search terms, not as your evidence. Find, open, evaluate, and cite the real sources.

Try: “Suggest search terms and likely source types. Do not fabricate citations or claim a source exists.”

5. Coding and data

Ask for explanations, tests, or debugging questions. Run code only in an approved environment and verify logic, dependencies, licenses, security, and outputs.

Try: “Explain this error and propose tests. Do not change files, install packages, or use credentials.”

6. Career preparation

Practice interviews or compare your experience with a job description. Never invent skills, achievements, employment, or credentials.

Try: “Ask behavioral interview questions based only on the experience I provide. Flag any unsupported claim.”

Trust needs evidence

AI can sound confident and still be wrong

Generated text may contain false claims, made-up citations, flawed calculations, insecure code, missing context, or biased assumptions. Verification is part of the work.

Claims and sources

  • Open every source; do not cite a link you have not read.
  • Confirm author, title, date, publisher, and supporting passage.
  • Prefer primary sources when possible.
  • Separate fact, interpretation, and uncertainty.

Math, data, and code

  • Redo important calculations.
  • Check units, assumptions, missing values, and sample limits.
  • Test normal, edge, and failure cases.
  • Do not run unknown code with sensitive files, credentials, or excessive permissions.

Meaning and fairness

  • Compare against course materials and context.
  • Look for stereotypes, missing perspectives, and unsupported certainty.
  • Ask what evidence would disprove the answer.
  • Escalate high-consequence questions to a qualified person.

Good verification is visible. Keep notes showing the source you checked, the error you found, the change you made, and the reasoning behind your final answer.

Pause before uploading

Protect privacy, accounts, and devices

Anything entered into an AI service may be stored or processed under that service’s terms and your institution’s agreement. Use an institution-approved account when provided and share only what the task requires.

Do not enter

  • Passwords, recovery codes, API keys, or security answers
  • Government IDs, financial, health, or accommodation information
  • Private messages or another person’s personal information
  • Exams, answer keys, confidential work, or proprietary data
  • Unpublished research or restricted course materials without permission

Safer account habits

  • Use a unique password and multi-factor authentication.
  • Review privacy, sharing, memory, history, and training settings.
  • Remove unnecessary chat history or connected data when allowed.
  • Do not approve unexpected login or tool-access requests.
  • Report a suspected exposure promptly.

Agents and connected tools

  • Check exactly what an AI tool can read, change, send, buy, or delete.
  • Grant the least access and shortest duration possible.
  • Preview actions before approval.
  • Do not connect school, email, drive, or financial accounts casually.
  • Know how to revoke access and undo changes.
Deepfakes and impersonation: treat urgent voice, video, text, or payment requests as unverified. Contact the person or institution through a known channel, never share a one-time code, and report suspected impersonation.

Own the final work

Before-you-submit checklist

Rule and authorship

  • I followed the AI rule for every step.
  • The final work reflects my own understanding and judgment.
  • I can explain the claims, methods, code, and choices.
  • I did not invent experience, data, results, or sources.

Evidence and quality

  • I opened and checked every citation.
  • I recalculated important numbers and tested code.
  • I checked quotations, paraphrases, permissions, and attribution.
  • I reviewed bias, missing context, accessibility, and tone.

Privacy and disclosure

  • I did not expose private or restricted information.
  • I used the required disclosure or citation format.
  • I kept any prompt, draft, or process record the instructor requires.
  • I asked before submitting if a rule remained unclear.

Simple disclosure model

“I used [tool and version, if known] on [date] to [permitted purpose]. I reviewed the output by [verification steps], changed [what you changed], and take responsibility for the submitted work.” Follow your instructor’s format if it differs.

Choose the right citation and disclosure

Permission, source checking, disclosure and citation answer different questions. Follow the rule for the task first. If AI helped you find a source, read it and cite the source you actually used. If you quote or analyze generated material, follow your required style’s AI guidance. Explain meaningful assistance separately when required.

APA and MLA use different reference patterns

These are patterns with placeholders, not references to actual chats. Replace every field with verified information and check the linked official guidance. For an APA tool reference, verify the tool’s update year; for a chat reference, use the conversation date.

  • APA, a tool: [Developer company]. ([Year]). [Tool or model name] [Large language model]. [Tool URL]
  • APA, a chat: [Developer company]. ([Year, Month Day]). [Chat title] [Generative AI chat]. [Tool or model name]. [Chat URL]
  • MLA, generated material: “[Description of generated material]” prompt. [Tool name], model [specific model], [Developer company], [Day Mon. Year], [stable conversation URL].

APA distinguishes a tool reference from a chat reference. MLA asks for the specific model and prefers a stable conversation link when available; its guidance gives a tool-link alternative when that feature is unavailable. Do not guess dates, model names or URLs. A citation does not authorize AI use or verify the answer.

Before sharing a conversation, check for private information, restricted materials and account details. If sharing is unsuitable, ask your instructor for an approved documentation method. A general tool reference, selected permitted excerpts or a description of the method may serve a different purpose; use the required style and local directions.

APA Style AI reference examples · MLA AI citation guidance · MLA guidance on describing AI use

Optional reflection: what did I do, and what did AI do?

Use this short record when it helps you learn or when your instructor requests it. Write down [skill I am practicing], [rule I checked], [work I will do myself], [permitted AI help], [source or test I used to check it], [a suggestion I changed or rejected and why], and [required disclosure]. Keep the minimum record needed under your institution’s privacy and storage rules.

Fictional community-college example: An instructor permits AI to ask practice questions but requires students to write their own argument. A student uses an approved assistant to question a claim, rejects a suggestion unsupported by the assigned reading, checks the relevant passage and writes the paragraph independently. Their disclosure explains the questioning help, the rejected suggestion and the source check. This illustrates a process; it reports no classroom result.

The exercise is a MAIN learning aid. A contribution statement is not an APA Style requirement unless an instructor or institution requires it. A self-report alone does not prove authorship or learning.

Cornell AI Critical Literacy Program · APA-hosted discussion of contribution statements

Continue with MAIN

Build your AI skills in Mississippi

Student resources

Find student-focused AI information, learning opportunities, and practical support.

Explore student resources

Courses and credentials

Review MAIN’s course catalog for current availability and registration steps.

Browse AI courses

Faculty AI Playbook

See the companion guidance instructors can use to set course rules and design meaningful assessments.

Open the Faculty AI Playbook

AI + Dual Enrollment Guide

Understand how college course expectations and K-12 technology rules should be coordinated.

Open the dual-enrollment guide

Student AI questions

Frequently asked questions

Is using AI cheating?

It depends on the course, assignment, and institution. Read the instructions and syllabus, follow the stated boundary, disclose or cite AI use when required, and ask the instructor before using AI if the rule is unclear.

Can I use AI to study?

When course rules allow it, AI can help explain a concept, generate practice questions, compare approaches, or give feedback. Check the output against course materials and use the tool to practice thinking rather than replace it.

What should I avoid uploading?

Do not upload passwords. Do not upload financial or health information, government identification numbers, private messages, confidential work, proprietary data, unpublished research, exams, answer keys, or another person’s information unless an authorized workflow explicitly allows that specific data use.

How should I check an AI answer?

Compare it with course materials and reliable sources, open every citation, redo calculations, test code in a safe environment, look for missing context or bias, and ask a qualified person when the consequence of error is high.

How should I disclose AI use?

Follow the instructor’s required format. A useful disclosure names the tool, explains what it helped with, identifies what you changed or verified, and distinguishes your own reasoning and final responsibility.

A detector score is not proof of misconduct. If your work is questioned, ask which rule applies and use drafts, sources, notes, or an explanation of your process. MAIN’s evidence guide to AI detection and verifying learning explains how fair review should work.

Use AI to strengthen the work only you can do

Your questions, evidence, judgment, creativity, and ability to explain a decision remain central. Keep learning visible and keep yourself accountable.

Search MAIN

Courses, guides, news, and information.