Skip to content
Createfirm

AI risk management for small business: a practical starter

A plain-English AI risk management starter for small business, built on the NIST AI Risk Management Framework, with a 12-step checklist for a small team.

By the Createfirm teamPublished 7 min read

Illustration of AI risk management: a steel shield with a gold star circled by four orbit rings carrying tool icons

AI risk management for a small business means four habits: decide who is accountable, list where you use AI and what could go wrong, test and monitor the results, and fix problems fast. That structure comes from the NIST AI Risk Management Framework, and a 20-person company can run it with a short checklist.

What is the NIST AI Risk Management Framework?

The NIST AI Risk Management Framework (AI RMF) is a guide from the National Institute of Standards and Technology, released on January 26, 2023. NIST describes it as intended for voluntary use.

The framework document (NIST AI 100-1) says it is meant to be “voluntary, rights-preserving, non-sector-specific, and use-case agnostic,” with flexibility for “organizations of all sizes.” It also notes that small to medium-sized organizations may face different challenges than large ones, depending on their capabilities and resources.

The framework describes trustworthy AI as valid and reliable, safe, secure and resilient, accountable and transparent, explainable and interpretable, privacy enhanced, and fair with harmful biases managed.

NIST’s page states that AI RMF 1.0 is being revised, so check the page for the current version when you set up your process.

What do Govern, Map, Measure and Manage mean?

The framework has four functions. Here is each one in plain terms.

Function What NIST says What it means for a small business
Govern A cross-cutting function that enables the other three. It covers policies, accountability and third-party risk. Name an owner and write the rules down.
Map Establishes the context to frame risks related to an AI system. List each AI use, who it touches and what could go wrong.
Measure Uses quantitative, qualitative or mixed methods to analyze, assess, benchmark and monitor AI risk. Test before launch and keep checking.
Manage Allocates risk resources to mapped and measured risks, with plans to respond to, recover from and communicate about incidents. Fix the biggest risks first and have an incident plan.

What risks does the NIST Generative AI Profile add?

In July 2024 NIST published the Generative AI Profile (NIST AI 600-1), a companion to the framework for tools such as chatbots and content generators. It defines 12 risks that are “unique to or exacerbated by” generative AI.

These are the ones a small business is most likely to meet first:

  • Confabulation. NIST defines this as “the production of confidently stated but erroneous or false content,” often called hallucinations.
  • Data privacy. Leakage or unauthorized use or disclosure of personal or sensitive data.
  • Information security. Lowered barriers for attacks such as phishing and malware.
  • Intellectual property. Eased production or replication of content that may be copyrighted, trademarked or licensed.
  • Human-AI configuration. People over-relying on a system or treating it as more human than it is.
  • Value chain and component integration. Third-party components that are not transparent or traceable, and improper supplier vetting.

The profile focuses on four areas: governance, content provenance, pre-deployment testing and incident disclosure. The checklist below follows the same emphasis.

What does the FTC expect from businesses using AI?

The Federal Trade Commission applies existing consumer protection law to AI. Announcing Operation AI Comply in September 2024, the agency said its actions “make clear that there is no AI exemption from the laws on the books.”

One case in that sweep is a useful lesson. The FTC alleged that a company promoted an AI service as a substitute for a human lawyer without testing whether its chatbot’s output matched that level. The takeaway: test before you make a claim, and keep the evidence.

For data, the FTC’s Start with Security guide for business draws lessons from the agency’s data security cases. Three apply directly to AI tools: do not collect personal information you do not need, control access to data sensibly, and make sure your service providers implement reasonable security measures.

A 12-step AI risk checklist for a 20-person company

This checklist is our practical reading of the four NIST functions for a small team. It is a starting point, and the framework itself goes much deeper.

Govern

  1. Name one accountable owner. One person answers for how the company uses AI. This reflects NIST’s call for accountability structures.
  2. Write a one-page AI use policy. Cover approved tools, what data may never be entered, and when a person must review output.
  3. Train everyone once, then yearly. Thirty minutes on the policy and on confabulation is enough to begin.
  4. Vet your vendors. Ask each AI vendor how it handles your data, security and updates. NIST’s Govern function includes policies for third-party software and data.

Map

  1. Keep an AI inventory. One spreadsheet: tool, purpose, owner, data used, who is affected. Include AI features inside software you already own.
  2. Rate each use. Mark it low, medium or high based on who could be harmed and how hard a mistake is to undo. Customer-facing and money-related uses rate higher.
  3. Note the rules that apply. Record any legal requirements for each use, such as AI disclosure laws for chatbots.

Measure

  1. Test before launch. Run real questions, edge cases and attempts to push the tool off topic. Save the results.
  2. Sample outputs on a schedule. Review a set of real conversations or documents each month and track errors and escalations.
  3. Collect feedback. Give staff and customers an easy way to flag a wrong or harmful answer.

Manage

  1. Fix the highest-rated risks first. Add human review, narrow the tool’s scope, or switch it off for that use.
  2. Write a short incident plan. State who gets called, how to pause the tool, how to tell affected customers, and how to record what happened.

How do you start this month?

Start with steps 1, 2 and 5. An owner, a one-page policy and an inventory take a few hours and make every later step easier.

Good data habits support all of this, so pair the inventory with the steps in how to prepare your business data for AI. If you are still deciding where AI fits, our AI readiness assessment questions help you choose a first use with manageable risk.

This article is general information, not legal advice, so confirm any legal obligations for your business with your own counsel.

The short version

The NIST AI Risk Management Framework gives you four jobs: govern, map, measure and manage. For a small business that means an owner, a short policy, an inventory, testing before and after launch, and an incident plan. Start small, write it down and review it on a schedule.

Frequently asked questions

What is the NIST AI Risk Management Framework?
It is a voluntary framework from the National Institute of Standards and Technology, released on January 26, 2023, to help organizations manage the risks of AI systems. It is organized around four functions called Govern, Map, Measure and Manage. NIST describes it as flexible for organizations of all sizes and in all sectors.
Does a small business have to follow the NIST AI RMF?
No. NIST states that the framework is intended to be voluntary. It is useful as a shared structure and vocabulary for thinking about AI risk, and NIST wrote it to be flexible for organizations of all sizes. A small business can adopt the four functions at a small scale, with a short policy, a tool inventory, basic testing and an incident plan.
What are the biggest generative AI risks for a small business?
NIST's Generative AI Profile lists 12 risks. The ones most small businesses meet first are confabulation, which is confidently stated but false content, data privacy, information security, intellectual property, and value chain and component integration, which covers reliance on third-party tools you cannot fully see into.
How often should a small business review its AI risks?
NIST says AI systems should be tested before deployment and regularly while in operation, and that risk resources should be allocated on a regular basis. A practical rhythm for a 20-person company is a short monthly check of sampled outputs and incidents, plus a fuller review each quarter and whenever a tool or use changes.

Sources

Every fact in this article was checked against these pages on November 14, 2026.

  1. AI Risk Management Framework, National Institute of Standards and Technology
  2. Artificial Intelligence Risk Management Framework (AI RMF 1.0), NIST AI 100-1, National Institute of Standards and Technology
  3. Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile, NIST AI 600-1, National Institute of Standards and Technology
  4. FTC Announces Crackdown on Deceptive AI Claims and Schemes, Federal Trade Commission
  5. Start with Security: A Guide for Business, Federal Trade Commission

Keep reading