CPA firms

Attackers aren't hacking AI. They're using your team's curiosity about it.

Microsoft says a single phishing campaign sent up to 100,000 emails in one day — every one of them dressed up as ChatGPT.

1 min readOriginally posted on LinkedIn

Microsoft says a single phishing campaign sent up to 100,000 emails in one day — every one of them dressed up as ChatGPT.

What struck me is that the attackers didn't break into any AI company. They just noticed your team is curious about AI right now, and curiosity makes people click.

The lures are mundane. A billing notice from an AI tool nobody remembers subscribing to. A "required plugin" download. A sign-in page that looks exactly right. And the better ones don't steal your password at all — they sit in the middle and take the session after you approve the sign-in prompt yourself. Microsoft says it is now disrupting more than 45,000 of those middleman attacks every month.

For a CPA firm, think about what one partner's mailbox actually holds: client tax documents, bank details, a year of correspondence, and enough context to send a very convincing wire request in your name. The cleanup isn't the IT work. It's the calls you make to clients — during filing season.

Three things that help more than any tool. Decide which AI apps your firm actually uses and tell your team out loud, so everything else looks wrong. Move to sign-in that can't be replayed. And stop letting staff install software straight from a browser download.

Worth a look before your next cyber renewal.

ShareLinkedInEmail

Want the next one in your inbox?

One practical Microsoft 365, Copilot, Azure, or security move per week — plain English, no pitch.

Subscribe via email

All editions