Your Employees Just Did Your AI Market Research


A CEO told me last week he had banned ChatGPT inside his company. Then, halfway through our meeting, his head of marketing mentioned she had used it that morning to draft three ad campaign positioning messages.

He laughed. She laughed. Nobody had a plan.

That conversation is becoming the new normal.

IDC just reported that 65% of employees are already using AI at work. About 39% of them are on free, unapproved tools. IBM’s 2025 Cost of a Data Breach Report adds the other half of the story: 1 in 5 companies has now had a breach traced back to “Shadow AI,” the unsanctioned tools everyone uses and nobody owns.

Most CEOs read those numbers and feel a knot in their stomach.

I’d push back on that instinct.

Shadow AI is a signal.

Treat it as one before you treat it as a security problem. Your employees just did your AI market research for you, for free. Every workflow they are quietly pushing into ChatGPT, every spreadsheet they are feeding into Claude, every email draft they are running through Gemini. That is a heatmap. They are showing you exactly where AI saves the most time inside your company. They have already done the pilot. You just have not been invited.

The leaders who treat Shadow AI as a discovery layer move faster than the ones still circulating their fourth draft of an acceptable-use policy.

Here is the move I walk CEOs through. Three steps. From panic to playbook.

  1. See. Run a usage audit. Anonymous, no consequences, no judgment. Ask: what AI tools are you using, for what tasks, with what data? Most companies are stunned by what comes back. One client found 17 different tools in use across a 200-person team. Another discovered that their best salesperson had built a custom GPT that was outperforming the official enablement deck. You cannot manage what you cannot see.
  2. Sanction. Pick two or three enterprise tools and stand them up properly. ChatGPT Enterprise, Claude for Work, Microsoft Copilot, whatever fits your stack and your data posture. Build a fast-track approval path for new tools so requests do not sit in a six-month queue while employees quietly default back to the free version. The goal is to give people a safe lane to run in.
  3. Scale. Take the workflows your audit surfaced and turn them into real pilots. Track real ROI. Hours saved. Errors caught. Deals closed. Make your best shadow users the teachers. Share wins quarterly. Tie one or two AI initiatives to a board-level metric so the work survives the next reorg.

Most CEOs skip step one because it feels uncomfortable to ask. The data will be messy. Some answers will surprise you. A few will make legal nervous.

But there are gems in the answers and they will help your org go faster with AI if you look for them.

Alex

Alex Talks AI

As an AI Coach, Advisor, and Agent Builder, I help organizations and business leaders harness the power of artificial intelligence to boost productivity and streamline operations. I enable organizations to navigate the transformative landscape of AI, educating teams, identifying operational and strategic opportunities with AI and creating a framework for safe and transparent use of data in the organization.

Read more from Alex Talks AI

I spent last week building an investment deck for a client. The raw material was a pile of research reports. The output needed to be a branded PowerPoint that looked like it came from inside their firm, not from a random consultant with a Canva account. If you've ever tried to get an LLM to spit out a polished, branded deck, you know how this usually goes. The content is fine. The formatting is a disaster. Here's what I tried. Attempt 1. I worked in Claude, pointed it at the folder of...

Their names are Aaron Sorkin, Andy Sachs, Hemingway, Darwin, Ted Lasso, and Archivist. They're agents I built inside Claude. Each one has a role, a personality, a set of files they own, and a clear job. Aaron Sorkin is my chief of staff. He directs everything. When I throw something into the void at 11pm, he decides whether it's an Andy problem, a Hemingway problem, or something I actually need to handle myself. Andy Sachs runs operations. She tracks my Notion CRM, drafts invoices, watches my...

You ask for research. You get a confident-sounding wall of text. The numbers feel right. The framing is fine. But you cannot quite tell where any of it came from, and you would not bet a client meeting on it. I had that feeling one too many times this month, so I ran a small experiment. Same research brief, different tools. The question: what is it actually like to work at SpaceX, xAI, and Tesla? I wanted real numbers from Glassdoor, Indeed, and Blind. Ratings, work-life balance, culture,...