Is using ChatGPT at work shadow AI?

Only if your organization hasn't approved it. Using an approved business account within your organization's rules isn't shadow AI; using a personal account for work information usually is.

How do I find out what AI tools my team uses?

Ask. An anonymous, no-blame survey is the fastest and most honest way. Technical monitoring can help larger organizations, but it works best alongside open conversation and approved alternatives.

Should every business and church worry about shadow AI?

Yes, in proportion. Teams of every size handle sensitive information, like customer details, donor records and prayer requests. A short policy and one approved tool cover most of the risk.

What is shadow AI, and how do you manage it?

By Rafael · Updated October 20, 2026 · 3 min read

The short answer

Shadow AI is people using AI tools at work that the organization hasn't approved, usually free personal accounts. It's common and rarely malicious, but company and customer information can leave your control. Bans tend to push it underground; the fix is approved tools, simple rules and training.

What shadow AI means

Shadow AI is the AI version of "shadow IT": staff using tools the organization doesn't know about or hasn't approved. Most often it's someone pasting an email, a spreadsheet or meeting notes into a free chatbot on their personal account to get work done faster.

It's widespread. In a 2026 PagerDuty survey of 1,250 office professionals at large companies, two-thirds (66%) said they had used AI tools at work they believed weren't permitted.

Why it's a risk

Data leaves your control: customer, donor, member or employee information ends up in accounts and systems you don't manage.
Training and retention: depending on the plan and settings, what's typed into personal AI accounts may be stored or used to improve the service.
No visibility: you can't support, secure or learn from tools you don't know are being used.
Quality: AI output that goes out unchecked can contain mistakes in your organization's name.

Why banning AI backfires

People use shadow AI because it helps them. A ban removes the safe options but not the need, so use continues quietly on personal phones and accounts, where you have the least protection. The goal is to bring AI into the open, not to stamp it out.

How to manage shadow AI: a practical plan

Ask, without blameRun a quick, anonymous survey: which AI tools do people use, and for what? Make it clear no one is in trouble. You'll learn what's genuinely helpful.
Approve good toolsGive people a business or team plan of a major AI assistant. Business plans generally come with stronger data commitments and admin controls than personal accounts.
Set simple rulesWrite a one-page AI policy: what's approved, what data never goes in, and what needs a human check. Our AI policy template is a starting point.
Train on real workAn hour of hands-on practice with the approved tool beats any memo. See how to train your team on AI.
Make asking easyName a go-to person for AI questions and new tool requests, and review the approved list every few months.

Copy-ready note to your team

Send this when you announce your approved AI tool. Adjust the brackets for your organization.

Team announcement: using AI safely

Subject: Using AI at work — what's approved and how to use it safely

Hi team,

Many of us already use AI tools to write, summarize and plan. That's great, and we want to make it easier and safer.

What to use
• For work, please use [approved tool and plan, e.g. "our company Claude/ChatGPT/Copilot business account"]. Ask [name] for access.
• Personal or free AI accounts are fine for personal things, but not for work information.

What never goes into any AI tool
• Passwords, keys or account numbers
• [Customer / donor / member] personal details, unless the task truly needs them and the tool is approved for it
• Health, HR, counseling or other confidential information

Good habits
• Remove names and identifying details when you can.
• Check facts and numbers before anything goes out.
• You're responsible for what you send or publish, even if AI drafted it.

Not sure? Ask [name]. No one is in trouble for asking, or for telling us which tools you've been using. We'd rather know, so we can get you a safe version of what helps.

Our full AI policy: [link]

Thanks,
[Name]

Common questions

Is using ChatGPT at work shadow AI?: Only if your organization hasn't approved it. Using an approved business account within your organization's rules isn't shadow AI; using a personal account for work information usually is.
How do I find out what AI tools my team uses?: Ask. An anonymous, no-blame survey is the fastest and most honest way. Technical monitoring can help larger organizations, but it works best alongside open conversation and approved alternatives.
Should every business and church worry about shadow AI?: Yes, in proportion. Teams of every size handle sensitive information, like customer details, donor records and prayer requests. A short policy and one approved tool cover most of the risk.