Which Companies Banned ChatGPT at Work and What to Use Instead

BLUF: A growing list of large employers has banned or tightly restricted ChatGPT at work, almost always to stop confidential data from leaving the company. Reported cases include Samsung, Apple, JPMorgan Chase, Bank of America, Citigroup, Goldman Sachs, Wells Fargo, Deutsche Bank, Verizon, Amazon, Accenture, and Northrop Grumman. If your employer is one of them, the safe path is a private or on-device AI tool that processes text locally or through an approved, contract-backed channel, not the public ChatGPT web app.

Which companies banned ChatGPT at work

The most cited example is Samsung, which restricted internal use in 2023 after employees pasted confidential source code into the public chatbot. Apple limited employee use of ChatGPT and similar tools around the same time. On Wall Street, JPMorgan Chase moved first, and Bank of America, Citigroup, Goldman Sachs, Wells Fargo, and Deutsche Bank followed with their own restrictions. Verizon, Amazon, Accenture, and the defense contractor Northrop Grumman have also limited or blocked access.

Why companies banned ChatGPT

The common thread is not that AI is unhelpful, it is that data handling is unclear. When you paste a contract, customer record, or unreleased plan into a public chatbot, that text leaves your network and may be stored or used to improve the service. For banks, healthcare, defense, and anyone under GDPR or HIPAA, that is a compliance problem before it is anything else. Companies that banned ChatGPT at work are usually managing legal and confidentiality risk, not making a judgment about the technology.

What "banned" actually means

A ban almost never means "no AI ever." In practice these policies block the free, public app while leaving room for approved alternatives: enterprise tools with signed data-processing agreements, internal models hosted on company infrastructure, or on-device software that never sends your text anywhere. Read your acceptable-use policy closely, because the line is usually drawn at where the data goes, not at the word "AI."

What to use instead

If you still need help understanding dense text, polishing a draft, or writing a clear reply, choose a tool that keeps data under your control. The safest category is on-device: a local LLM for writing on a Mac runs the model on your own machine, so nothing you select is uploaded. The next safest is an approved enterprise account where your employer has a contract covering data retention. A good private AI writing assistant that works offline can give you most of ChatGPT's everyday usefulness without the privacy exposure that triggered the bans.

How a private, on-device assistant fits

This is the gap Nugumi is built for. It is a Mac menu-bar assistant that can explain selected text in plain words, rewrite your own draft to read clearly and professionally, and draft a reply in your style, in any app with selectable text. The key detail for regulated workplaces is the backend choice: point it at local Ollama and it runs entirely on-device, so nothing you select leaves your computer. You can also use your own API key or a ChatGPT or Claude subscription if your employer approves that route. For teams that need a true offline AI writing app with no internet dependency, the local-model option keeps work inside the machine. You can download it and try the local setup before bringing it to IT.

What to check before using any AI tool at work

Before you paste anything sensitive, confirm three things. First, read your company's acceptable-use or data policy and note what is explicitly forbidden. Second, find out where the tool processes text and whether it retains it; on-device tools that keep data local are the easiest to clear. Third, when in doubt, ask IT or security rather than assuming. A short message now is cheaper than an accidental leak later, and most teams will gladly approve a tool that demonstrably keeps data on your own machine.

The bottom line

Bans on ChatGPT at work are about data leaving the building, not about whether AI is useful. The companies on this list still let employees use AI, they just require it to be private, approved, or on-device. Pick a tool that matches that bar, and you keep the speed of an AI assistant without the exposure that got the public app blocked in the first place.