AI ParametersWorkplace AI guardrails
Get the Guardrails Toolkit — $299

60-day money-back guarantee. If it doesn't work for your team, you get your money back.

Menu

AI WORKPLACE GUARDRAILS

Your team is already using AIHave you set the parameters?

Answer a few guided questions. Get your company's AI rulebook — policy, training, rollout kit. No consultants. No per-seat fees.

Get the Guardrails Toolkit — $299See a policy and training example

60-day money-back guarantee. If it doesn't work for your team, you get your money back.

Get a free sample control card

Your email is saved with your request. Privacy

One company licenseNo per-seat fees
No employee accountsShare the training your way
No sensitive uploadsYour materials stay with your company
Built for self-serviceNo consulting engagement required

THE PROBLEM

Good employees are trying to work faster

Good employees take shortcuts to work faster. Nobody told them where the line is.

1.

Nobody said what goes in the prompt

Before the toolkit

Your team uses an approved AI tool to prepare client proposals. But nobody has clarified whether client notes, pricing, or contract details belong in the prompt.

After the toolkit

Write down once what can go into the AI — and what can't.

2.

The AI promised a delivery date we never agreed to

Before the toolkit

AI helps a salesperson draft a proposal. It reads well, but includes a delivery date the operations team hasn’t agreed to.

After the toolkit

Set the rule: pricing, delivery dates, and promises get a human check before they go out.

3.

Nobody decided what the AI is allowed to do

Before the toolkit

A team connects an AI assistant to company email and files. Nobody has clearly decided whether it can only read and draft—or also send messages and change records.

After the toolkit

Write down what the AI can touch, who owns it, and when a human must approve.

WHAT YOU RECEIVE

Answer once. Everything matches.

The policy, training, and rollout kit all come from the same answers — so they all say the same thing. No more folder of mismatched templates.

1.

Company AI policy

Start with a ‘Before you approve a tool’ checklist, then set clear rules for tools, company information, and human review.

2.

Employee training

Give employees a practical lesson and a simple way to acknowledge completion.

3.

Manager rollout kit

Launch with ready-to-use emails, talking points, team prompts, and new-hire steps.

4.

Control cards and records

A written record of every AI tool, who approved it, and what it's allowed to touch.

See a policy and training exampleSee the complete toolkit

INCLUDED AI USE CONTROL CARD

Can your AI draft an email, or send it too?

Record what a connected AI tool can access and do, who’s responsible for it, and how to check its limits.

The card doesn't set or enforce software permissions. Your team still configures the tool.

See the control card
COMPLETED EXAMPLESee what done looks like
ChatGPT · client-facing writingSelect to read
COMPLETED EXAMPLE

AI USE CONTROL CARD — SAMPLE

Free sample from AI Parameters — setparameters.com. The full toolkit has one of these for every AI tool your team touches.

Why this exists: A staffer drafts a client email in ChatGPT and pastes in the client's name, account number, and private details "to get the tone right." That data now lives on someone else's servers, outside anything you control — and nobody told them not to, because the rule was never written down. This card is that rule.

The rule in one sentence: Use AI to write better. Never feed it anything you wouldn't tape to your front window.

Tool: ChatGPT (any version) — drafting client emails, marketing copy, summarizing documents

Who may use it: Anyone on the team whose job includes writing to clients or the public

This tool MAY see:

  • Your own draft words (you type them in)
  • Public facts about the business — address, hours, services already on your website
  • Anonymous examples — "a customer complained about a late delivery" (no names, no numbers)

This tool must NEVER see:

  • Client names, phone numbers, emails, or addresses
  • Account numbers, policy numbers, financial figures, SSNs
  • Employee personal info
  • Anything marked private or confidential

Before anything AI-written goes to a client or the public: the person who wrote the prompt reads the output word for word and signs off on it. No exceptions, no "it looked fine."

The 30-second check (all three, every time):

  1. Facts are right — names, dates, prices
  2. No client info slipped into the prompt or the output
  3. You'd be comfortable if the client saw the exact words you typed into the AI

If someone breaks this: tell the owner or manager the same day, and the rule gets updated so it can't happen the same way twice.

How to use this card: Print it. Hand it to everyone who writes to clients or the public. Takes five minutes to explain. That's the whole rollout.

This is part 1 of the card. Part 2 covers who approved it, how you test the limits, and what to do when something goes wrong.

Your team doesn't use one AI tool. They use five. This is the card for one of them. The full AI Workplace Guardrails toolkit gives you both parts of this card for every tool they touch — fill one out per tool — plus the company policy, the employee training, and the manager rollout kit. $299, one time, yours forever.

WHERE GUARDRAILS FIT

Built around how your company uses AI

It's not about your industry. It's about what your team does with AI. A regulated company can still use guardrails for appropriate everyday work.

If your team uses AI to write, sell, or serve customers, this is for you.

You make the calls. The toolkit gives you everything to make them stick.

ONGOING GUARDRAIL UPDATES

A policy shouldn’t sit untouched for years

Every month, we tell you what changed in AI, what it means for your team, and which wording to update. You stay in charge.

See what the monthly service includes →

THE COMPLETE GUARDRAILS TOOLKIT

$299one-time toolkit purchase

Includes setup access, company materials, employee training, rollout tools, and the first 30 days of ongoing updates.

Then $39.99/monthAutomatic renewal. Cancel anytime.

Cancel anytime. Keep what you've got.

Get the Guardrails Toolkit

60-day money-back guarantee. If it doesn't work for your team, you get your money back.

Review pricing and FAQs →

Our approach is informed by practical risk-management and AI-literacy guidance, including the NIST AI Risk Management Framework and the U.S. Department of Labor AI Literacy Framework.