Manage
A friend recently told me a story that stopped me mid-conversation. His son landed a summer internship, and on day one, his manager handed him a task: go to a government website, download weekly data files, and compile everything against a specific set of parameters. The manager’s parting words? “This is your summer project.”
Thirty days of work. Handed to the new intern. Classic.
What the manager didn’t anticipate was that this particular intern knew how to use AI productivity tools for business in ways most of his colleagues didn’t. Without missing a beat, the intern used the Claude Chrome extension to pull all the data and Claude Cowork to compile and analyze it against the required parameters. Two days later, he walked into his manager’s office and presented the finished project.
The manager was floored. He genuinely believed it would take the entire summer.
There are two very important lessons in this story — one about opportunity, and one about risk. Every business owner needs to hear both.
The Productivity Gap Is Already in Your Building
Here’s an uncomfortable truth: the gap between how fast your team could work and how fast they actually work is probably enormous — and your interns already know it.
The intern in this story didn’t do anything magical. He used tools that are freely available and increasingly easy to use. What he did differently was know they existed and know how to apply them to a real problem.
Meanwhile, the rest of the organization was operating on the assumption that this task required 30 days of manual effort. Nobody stopped to ask: Is there a better way to do this?
That question — simple as it sounds — is one most organizations never ask out loud. We get locked into workflows that made sense five years ago. We assign repetitive tasks without questioning whether those tasks could be automated or dramatically accelerated. And we often miss the opportunity because the people most likely to know a smarter path (younger employees, newer hires) aren’t in a position to challenge the process.
The lesson here isn’t that interns are smarter than their managers. It’s that AI tools for employees have matured to the point where a single person, working alone, can compress weeks of repetitive data work into hours — and your team may already be doing this without any formal guidance in place.
What Happened Here Was Actually a Near Miss
Now for the part of the story that should keep you up at night.
The intern’s use of AI tools worked out fine in this case, because the data he was working with came from a public government website. It was non-proprietary, non-sensitive, and freely available to anyone with a browser.
Your company may not be that lucky.
When employees use AI tools — whether that’s Claude, ChatGPT, Copilot, or any other platform — they are often uploading data to cloud-based systems. In many cases, that data can be used to train models, stored on third-party servers, or exposed to terms of service that your legal or compliance team has never reviewed.
Think about what your employees work with on a daily basis:
- Client records and contact lists
- Financial reports and projections
- Contracts, NDAs, and legal documents
- Patient data, if you’re in healthcare
- Proprietary pricing, formulas, or processes
Now ask yourself: do you have a policy in place that prevents any of that from being uploaded to an AI tool without authorization? If you don’t, there’s a real chance it’s already happening.
The intern in this story got lucky. The data was public. But the behavior — uploading files to an AI platform without checking with IT first — is exactly the kind of well-intentioned action that can result in a serious data breach, a compliance violation, or a very expensive conversation with your attorney.
The Right Response Isn’t to Ban AI — It’s to Get Ahead of It
Some businesses react to stories like this by wanting to lock everything down. Block the AI tools. Restrict access. Issue a memo.
That’s the wrong instinct.
The intern in this story showed you something valuable: your team can move dramatically faster when they’re equipped with the right tools. The goal isn’t to suppress that capability — it’s to channel it safely.
Here’s what a smart response actually looks like:
1. Ask Your Team What They’re Already Using
Before you write a policy, find out what’s already happening. Ask your employees — especially your newer ones — whether they’ve used AI tools to get their work done. You may be surprised by what you find. More often than not, people are already experimenting, and they’ll tell you what’s working if you create space to ask.
2. Build an Acceptable Use Policy with IT
Work with your IT team or managed service provider to define which AI tools are approved, which categories of data can and cannot be uploaded, and the review process before a new tool is used for company work. This doesn’t need to be a 30-page document. A clear one-pager that employees actually read is worth more than a policy binder nobody opens.
3. Separate Public Data from Proprietary Data
Not all data carries the same risk. Help your team understand the difference between publicly available information (like that government dataset) and confidential or regulated information. A simple classification framework — even just “green / yellow / red” — gives employees a mental model they can apply in real time without needing to escalate every decision.
4. Make IT the Partner, Not the Gatekeeper
The fastest way to ensure employees work around your security policies is to make IT feel like an obstacle. The better model is to position your IT team or MSP as the people who help you use AI tools safely — not the people who say no to everything. When employees know they can bring a new tool to IT and get a thoughtful answer, they’re far more likely to ask before uploading sensitive files.
If You Don’t Have an IT Department, You’re Not Off the Hook
Many small businesses don’t have in-house IT. If that’s your situation, the absence of an IT team doesn’t reduce your exposure — it increases it.
Without someone responsible for reviewing tools, enforcing data policies, and monitoring for unusual activity, you’re relying entirely on your employees’ individual judgment. And individual judgment, however well-intentioned, isn’t a security program.
This is exactly where a managed service provider (MSP) adds value beyond just fixing computers and managing email. A good MSP helps you:
- Evaluate which AI tools are safe to use in your environment
- Build lightweight data handling policies your team will actually follow
- Ensure any AI-powered workflows meet your compliance obligations — especially if you’re in healthcare, legal, finance, or accounting
- Protect you from the kind of accidental data exposure that can turn a productivity win into a regulatory headache
At eMDTec, we work with small businesses and professional services firms across New Jersey every day on exactly these challenges. We’re not here to slow your team down — we’re here to help you move faster without creating risk you don’t know you have.
The Bottom Line: Your Interns Might Be Ahead of You
The story of that intern finishing a 30-day project in two days isn’t just a fun anecdote. It’s a signal.
The tools are here. Your employees — especially your younger ones — are already using them or ready to. The question isn’t whether AI productivity tools will change how work gets done in your organization. They already are.
The only question is whether you’re steering that change or just reacting to it after something goes wrong.
Ask your team what they’re using. Talk to your IT partner about which guardrails should be in place. And if you don’t have that IT partner yet, this is a good moment to find one.
Ready to assess your AI readiness and make sure your team is working smarter — without the data risk? Reach out to eMDTec and let’s have a conversation.