So, picture this: I open up Microsoft Edge this morning—you know, that browser we all use when we've given up on life—and there it is, staring me in the face like an unwanted relative at a family reunion: "Ask Copilot." Right smack in the middle of the PDF viewer. Curiosity gets the better of me, and I click on it. Suddenly, Bing Chat pops up, Microsoft's very own AI chatbot, eager to assist. I think, "Why not put this to the test?" So I ask it to summarize a long whitepaper I'm reading. After a few moments of digital digestion, it spits out a detailed—albeit incomplete—summary.

‍

Now, none of this is shocking. Microsoft has been on a mission lately: "Copilot everything!" It's like Oprah handing out cars, but instead, it's AI in every corner of their ecosystem. Windows 11, Office 365, Microsoft Edge—you name it, they're shoving Copilot into it. And let's be honest, with over a billion users across these platforms, this thing is spreading faster than gossip in a Cairo café. People are using it for everything: summarizing documents, planning strategies, crafting marketing fluff—you know, the usual corporate jazz.

‍

I read somewhere that real estate agents in Australia are using ChatGPT to manage customer relationships. Imagine that—your dream home recommended by a robot! Meanwhile, Salesforce isn't sitting idle. They're weaving AI chatbots into their products faster than you can say "upgrade available," ready to unleash them on salespeople worldwide.

‍

But here's the kicker: we have no clue how people are actually using these AI chatbots. It's like everyone joined a secret club, and the first rule is "don't talk about AI Club." No data, just anecdotes. These tools are popping up everywhere, and people are using them daily, but mum's the word on the specifics.

‍

Why the silence? Maybe because there's no clear policy. Employees are dabbling with AI but aren't sure if the boss approves. No guidelines, no best practices—just a lot of grey areas. It's like everyone is secretly dating the AI chatbot but doesn't want HR to find out. And this hush-hush usage could be weaving AI errors into the fabric of businesses everywhere. If these chatbots were perfect little angels, this wouldn't be a problem. But let's face it—they're not.

‍

Just today, I stumbled upon a post detailing how someone could launch a "prompt injection" attack via a sneaky email to a Gmail account monitored by Google Bard. Sounds complicated, but think of it as whispering bad advice into the chatbot's ear. Researchers have been warning about this since May 2023, but now we're seeing step-by-step guides. With Google Bard reaching over a billion users, that's a pretty massive playground for mischief.

‍

And let's talk about accuracy—or the lack thereof. These AI chatbots can be about as reliable as a street vendor's watch. They generate responses that feel true but might be as factual as a unicorn sighting. Sure, they've got access to the web now, but the internet is a mixed bag of truths, half-truths, and complete nonsense. Plus, the web can be another avenue for those prompt injection attacks. So, using the internet as their fact-checker is like asking a compulsive liar to verify your alibi.

‍

So here we are: AI chatbots are infiltrating organizations, expanding the threat landscape, and possibly spreading misinformation like it's confetti at a parade. Telling everyone to stop using them would be like trying to put toothpaste back in the tube—we don't even know who's using them! But we can start by taking a good, hard look at our own organizations. Let's find out where and how these chatbots are being used. It's not a fix-all solution, but understanding the scope of the issue is a step in the right direction.

‍

We have to start somewhere, right?

‍

It begins with open, judgment-free conversations. Chat with your team members casually about their AI usage. They'll only open up if they don't fear repercussions. Storming in with strict policies and mandates will just push the activity underground until something goes horribly wrong, and then it's an even bigger mess. Gentle and broad inquiries are the way to go. Ask the quiet questions to learn how AI is already part of your organization's daily life.

‍

And hey, if you need a hand starting these conversations—or figuring out how to safely and wisely implement AI in your organization—don't hesitate to reach out!