ZDNET’s key takeaways
- There’s a new scam involving ChatGPT.
- The scam starts with a sponsored Google link.
- The link leads to a custom GPT that gives malicious information.
A new scam involving ChatGPT has emerged, and it might lead you to install malware on your computer if you’re not careful.
Over the past several days, reports have surfaced of users encountering a strange interaction when using ChatGPT. “Service Availability Notice,” the message reads. “We’re currently experiencing limited availability on the primary domain. Please choose one of the following options to continue:”
Also:This new cyberattack tricks you into hacking yourself. Here’s how to spot it
The chat presents an offer to upgrade to a Plus subscription or a link to a backup domain. Here’s how the scam works, why it’s so convincing, and what you can do to avoid it.
(Disclosure: Ziff Davis, ZDNET’s parent company, filed an April 2025 lawsuit against OpenAI, alleging it infringed Ziff Davis’ copyrights in training and operating its AI systems.)

Artie Beaty/ZDNET
The custom GPT scam
The scam starts when you search “ChatGPT” in Google. Like almost every Google search, you’ll see sponsored results, or paid ads, above the authentic search results. If you click on what appears to be a real link to ChatGPT, you’re taken to what appears to be the actual site.
Instead, you’ve landed on a custom GPT, or a user-generated, specialized version of ChatGPT that’s tailored for specific tasks. In this instance, the task is to spit out the availability notice response no matter what’s entered.
What’s most convincing is that you’re on the actual ChatGPT domain and your account is logged in (if you were before). There’s nothing at first to indicate this isn’t authentic other than the name “Plus 5.6.” If you’re not familiar with ChatGPT model naming conventions, though, this wouldn’t stick out.
Also:I’m a tech pro, and an AI job scam almost fooled me
If you click the link in the response (it’s a free site hosted on Google Sites and obviously not a ChatGPT domain), you’re presented with a fake Cloudflare verification and instructions to paste and run a command in Windows/PowerShell. Running that command installs malware on your computer.
I tried the process myself and clicked on the first result after a simple search for ChatGPT. I landed on one of the scams. No matter what I typed, I got the same “limited” response with a link. My editor here at ZDNET tried the same process and got the normal ChatGPT, so it doesn’t appear to be all sponsored results.
Google told me it had deactivated several accounts associated with the specific ad campaign I saw, but that, of course, doesn’t mean this issue won’t pop up again, so it’s good to be aware.
How to stay safe
Here’s how you can avoid this specific scam, and some general advice for staying safe online.
- Type chatgpt.com straight into your browser instead of searching for it.
- Avoid clicking on (or at least be highly suspicious of) sponsored links in Google. These ads have a history of being malicious, so much so that Google has even deployed Gemini to detect and block bad ads.
- Never paste and run a command on your computer if you’re not certain what it will do, especially if a link or pop-up tells you to.
Roman Oliinyk, CEO and founder of PayCore Media Inc, a network security specialist who has spent 10 years building data-leak protection systems for large US companies, noted that a real Cloudflare check would never ask you to do anything on your keyboard. “At most,” he said, “it asks you to check a box or press a button.”
Also: This simple ChatGPT trick helps you spot scams before you click or respond
Oliinyk also recommended treating sponsored results as ads (which they are) and treating a link from a chatbot as you would a link from a stranger.
I’ve reached out to both Google and OpenAI for comment. OpenAI hasn’t responded yet, but a Google spokesperson told me: “Malvertising has no place on Google. We’ve suspended several advertiser accounts to block this campaign and we continuously update our defenses to stop new threats.”
