Hype and novelty
New features appear weekly, so a "new model" or "early access" message does not seem odd.
This short course is required for everyone at Charitybuzz. It takes about ten minutes: a few focused lessons, a quick quiz, and a confirmation that goes on your training record. Everything in it comes from the kinds of messages and requests our team is actually receiving, so what you learn here is what you will use at work.
AI tools are now part of daily work, which makes them a new favorite disguise. Scammers send fake messages "from" Anthropic, OpenAI, Google or Microsoft about billing problems, suspended accounts, usage limits, security incidents or brand-new features — all with a link to a copy of the sign-in page.
There is a second risk beyond fake emails: fake AI apps and browser extensions that promise a free or "pro" version and quietly steal what you type, and the habit of pasting company or partner data into tools we have not approved. This module covers both.
A fake AI billing notice aimed at the person who manages the company's AI workspace. Click every detail that should make you stop. There are 7.
Dear Customer,
We were unable to process the latest payment for your organization's workspace. Access for all members will be suspended in 12 hours and all projects and chat history will be permanently deleted.
To speed up verification, paste your organization API key into the form.
Quick check: for each item, decide whether it is a red flag or looks fine. Real addresses: Claude is at claude.ai (company site anthropic.com), ChatGPT at chatgpt.com (company site openai.com).
https://claude.ai/This is the real Claude address. Typing it yourself is the safest way to reach it.
https://claude-ai-login.com/billingA look-alike with extra words. The real owner of this address is claude-ai-login.com, which is not Anthropic.
https://anthropic.billing-support.net/update-paymentThe real owner is billing-support.net. "anthropic" is just a name added in front.
An email: "Your AI subscription is suspended. Update your card within 24 hours or lose all conversations"Threats to delete your data and a short deadline are pressure tactics. Check billing by going to the tool yourself.
A browser extension that promises "free unlimited AI" and asks to read and change all your data on all sitesAn extension with access to every site can read your email and files. "Free pro" offers are a common lure.
Opening the AI tool your company approved by typing its address yourselfGoing to a tool yourself is always safer than following a message.
New features appear weekly, so a "new model" or "early access" message does not seem odd.
People pay for these tools and rely on them, so a billing alert feels urgent.
API keys and logins can be sold or used to run up bills and reach company data.
Because the tools are new, we are less sure what normal looks like.
Here is how AI-related risks show up at a company like Charitybuzz:
An email says your Claude, ChatGPT or Copilot account has a payment problem or has been suspended and asks you to sign in or update a card.
A message claims your API key has leaked or will be deleted unless you sign in and rotate it through a link.
Someone pastes a bidder list, donor details, consignment contracts or login details into a free AI tool or extension to "save time".
Hyphens and extra words such as "login" and "billing" are common in fake addresses.
Real billing issues can be seen inside your account.
Check the provider's console yourself.
Used to entice you to install or sign in.
They can read everything you see in the browser.
No real provider does so by message.
Granting access can expose your email and files.
If you would not email it to a stranger, do not paste it into an unapproved tool.
Type claude.ai, chatgpt.com or the tool's own address, and look at billing or notices there.
Never follow a message link to "rotate" a key.
If you are not sure a tool is approved, ask John before pasting anything.
Browser extensions and desktop apps with broad access need a check first.
Use Gmail's Report phishing and tell John.
Also turn on two-step verification for the tool, and sign out of other sessions.
If a key was shared, it should be revoked and replaced. Do it quickly.
In Gmail, open the message, click the ⋮ menu and choose Report phishing. Then email [email protected] with a one-line summary of what you did (clicked, opened a file, typed a password, approved a prompt). Fast reports let IT block the sender for everyone.
From a device you trust, go to myaccount.google.com by typing it yourself, change your password, and review Security → Your devices. If you reuse that password anywhere else, change it there too.
This is the most urgent case: the attacker may already be in your account. Contact Charitybuzz IT ([email protected]) right away so they can end every session and reset access.
Don't keep working on that device and don't try to "clean it up" yourself. Leave it on, disconnect it from Wi‑Fi if you can, and contact IT.
This creates the completion record Charitybuzz keeps for compliance. Please use your work details.
Your completion has been recorded. Thank you for helping keep Charitybuzz, our nonprofit partners, and our bidders safe.
Certificate ID: —
View & print certificate