Home · Business · IT, Data & Security · Cybersecurity

PARTLY

As of 13 August 2026, AI can only partly set up multi-factor authentication for your business.

This still needs a person who signs their name to it.

Can you do it?

5 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0

Skill neededchat-fluent

Who has to check ita colleague

What the alternative costsNo priced alternative is supplied in the available tool data; GitHub Copilot is described as an AI pair programmer, which is relevant only if a developer is adding MFA to custom software.

If this goes wrong: staff can be locked out, recovery methods can be exposed or an account can remain unprotected while appearing to be covered.

What to actually do

  1. Hand it to a person

    The route this page recommends

    A person who owns the outcome does this end to end, worth it when the failure is dear.

  2. Use a tool built for this

    Second choice
  3. Do it yourself

    The distant third

    A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

    How to actually do it

    1. Open the official administrator documentation for your identity provider and list the provider name, applications, user groups, current MFA settings and available second factors without copying any passwords or codes.
    2. Create or confirm two separate admin accounts and document a recovery route that does not depend on the same device or factor as the main account, without placing any secret in the chatbot.
    3. Paste the prepared details and the copyable prompt into a chatbot, then ask it to produce instructions only for the named provider and to identify any missing information before giving rollout steps.
    4. In the provider's admin console, compare each proposed menu and setting with the current official documentation, then configure a pilot policy for a small group rather than enforcing MFA for everyone immediately.
    5. Enrol the pilot users and test normal sign-in, a failed second-factor attempt, a new device, lost-device recovery, admin access, integrations and any documented emergency account while recording the outcome.
    6. Compare the test results with the AI-generated checklist and correct any missing exclusions, recovery routes or unprotected accounts before extending the policy to the remaining users.
    7. Send the drafted user message through your normal business channel, then enforce MFA in stages and keep the confirmed recovery procedure with your internal IT records.

    Prompt

    Help me plan and carry out a safe MFA rollout for a UK business. Do not ask for or accept passwords, one-time codes, recovery codes, private keys, customer data or screenshots containing secrets. Work from these details only:
    
    Identity provider or admin service: [name]
    Applications that need MFA: [names]
    Number and type of user groups: [groups]
    Available second factors: [authenticator app, security key, passkey, SMS or other]
    Devices users have: [devices]
    Current MFA status: [what is already enabled]
    Business constraints: [working hours, remote staff, shared accounts, accessibility needs]
    Existing recovery and break-glass arrangements: [description, without secrets]
    
    Give me:
    1. A provider-specific sequence of admin-console steps, clearly separating settings that enforce MFA from settings that merely allow it.
    2. A pilot plan for a small user group, including a test account and a documented rollback or recovery route.
    3. A checklist covering enrolment, sign-in, lost-device recovery, admin accounts, service accounts, integrations, exclusions and emergency access.
    4. A list of checks I must perform in the live admin console because you cannot see it.
    5. A short user message explaining the change in plain UK English.
    6. Stop and ask a question whenever the provider, setting or consequence is uncertain. Never invent menu names or claim that MFA is enabled unless I have confirmed it.

    Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

What it gets wrong

  • AI cannot see whether the setting in your live admin console actually enforces MFA or only permits users to enrol.
  • AI cannot safely handle passwords, one-time codes, recovery codes, private keys or other secrets on your behalf.
  • AI cannot know which legacy integrations, service accounts or shared accounts will fail after enforcement without your testing them.
  • AI cannot provide an independent assurance that every administrator and critical application is protected.
  • AI cannot carry the consequences of locking out your staff or leaving an account unprotected.

What caps this at PARTLY: private data access, verification cost and stakes of error.

How we scored this

Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.

AxisScore (0–2)
Output2
Inputs1
Verification1
Liability1
Effort delta2
Total7 / 10

FAQ

Can ChatGPT set up MFA for my business?
It can produce provider-specific instructions, rollout plans and troubleshooting checklists, but it cannot access your admin console or make the changes for you. An administrator must apply the settings, protect the recovery routes and test the result.
What information do I need to set up MFA?
You need the identity provider, applications, user groups, available second factors, user devices, current MFA settings and recovery arrangements. Do not paste passwords, one-time codes, recovery codes or private keys into an AI tool.
Can AI check that MFA is working?
AI can give you a test checklist, but it cannot independently confirm your live configuration. Test sign-in, failed-factor handling, lost-device recovery, administrator access, integrations and emergency access with a colleague before enforcing the policy widely.
What is the safest way to roll out MFA to staff?
Start with a small pilot group, keep a documented recovery route and test the applications and accounts that could be disrupted. Extend enforcement in stages only after the pilot results match the provider's documented settings.

Nearby answers

Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.

The newsletter

AI news, new answers and product picks, straight to your inbox.