Security & Data Governance

AI Policy for a Company in Malaysia: What It Is and a Template

What a company AI policy is, what a Malaysian SME should put in it, and a one-page fill-in template covering tools, data, review, incidents and PDPA.

  • By
  • Updated
  • 3 min read
  • Facts checked
Short answer

A company AI policy is a short written rule set that says which AI tools staff may use, what data may go into them, who checks the output, and what to do when something goes wrong. For a Malaysian SME, one to two pages is enough if it also reflects the PDPA.

Part of the guide: PDPA and AI in Malaysia: What Companies Must Know (2026 Guide)

On this page
  1. What is an AI policy for a company?
  2. What should a Malaysian SME's policy contain?
  3. What is the template?
  4. How do you roll it out?
  5. What next?
  6. Sources
  7. Frequently asked questions

A policy turns a vague worry ("are people using ChatGPT?") into rules people can follow. This page gives you the structure and a template you can copy.

Part of the guide: PDPA and AI in Malaysia. Checked 22 October 2026.

What is an AI policy for a company?

It is a written statement of how the company uses AI tools: approved tools, allowed and forbidden uses, data rules, review duties and incident handling. It sits alongside your IT policy, your privacy notice and your employee handbook. It is different from an employee-facing "do and don't" sheet, which we cover in the employee AI usage policy template.

What should a Malaysian SME's policy contain?

Section What to write
Purpose and scope Who and what it covers (staff, contractors, all AI tools, including AI inside other software)
Approved tools Named tools, plan type, account owner
Data rules Data classes and which may go into which tool
Personal data Link to PDPA duties, privacy notice, cross-border check
Human review What must be checked by a person before use
Prohibited uses Examples you decide in advance
Incident reporting Who to tell, how fast
Ownership and review Policy owner, review date

What is the template?

Copy this, fill the brackets, and have management and your lawyer review it.

  1. Purpose. [Company] uses AI tools to work faster and better. This policy protects customers, staff and the business.
  2. Approved tools. Staff may use only these tools for work: [Tool A, plan], [Tool B, plan]. Use company accounts only. Tool owner: [name].
  3. Data we never enter. [List: IC numbers, health data, payroll, client contracts, passwords, API keys, unpublished financials.] Exceptions need written approval from [role].
  4. Personal data. Enter personal data about customers or staff only into tools marked "personal data approved" in the list. Where possible, replace names and numbers with placeholders first.
  5. Checking outputs. A person checks any AI output that goes to a customer, supports a decision about a person, or contains figures, legal text or medical content.
  6. Do not. Use AI to make final decisions about hiring, firing, pay or credit without human review; to impersonate a person; or to copy confidential material into unapproved tools.
  7. Mistakes. Report any suspected leak, wrong output sent out, or unapproved tool use to [role, phone/WhatsApp] within [24 hours]. We would rather hear early than late. Suspected personal data breaches are escalated at once because notification to the Commissioner can be due within 72 hours.
  8. Review. The owner reviews this policy every [6] months and when a new tool or law appears.
  9. Acknowledgement. Each staff member signs that they read it.

How do you roll it out?

Name the policy owner and tools against the current vendor terms; for example, Anthropic publishes whether commercial and consumer plans are used for training. Then run a 45-minute session with real examples from the company's own work. A policy that is emailed and forgotten changes nothing. If you want it delivered as a staff session, see our coaching and courses, and the guide to HRD Corp claimable AI training for funding rules.

Leadership teams draft the governance rules for a policy like this during AI Strategy for Leaders, and staff finish their own version in Generative AI at Work.

What next?

Read What makes a good AI policy to check your draft, or book a call and we will review it with you.

Sources

(checked November 2026)

The template is general information, not legal advice. Have a lawyer adapt it to your business.

Frequently asked questions

Is an AI policy required by law in Malaysia?

No law requires a document called an AI policy today. The PDPA does require you to protect personal data, and a written policy is the practical way to show you did. A proposed AI Governance Bill may add duties later.

How long should an AI policy be?

One to two pages for most SMEs. A policy that fits on a page gets read. Put detail in appendices, such as the approved tool list.

Who should approve the policy?

Management, with input from whoever handles data protection (the DPO if you have one) and a lawyer for the clauses that carry legal weight.