Every organization is currently investing money in AI, and almost none of them know what their employees are actually doing with it”: Study finds that employees are using their personal AI accounts at work, raising a range of issues.

Two-thirds of AI usage on personal accounts is actually for business purposes
Workers also use company-provided tools to ask their personal questions.
Clumsy enterprise authentication makes trusted tools harder to access in an instant

A new study from Harmonic has claimed that almost two-thirds (64.5%) of all activity on personal and free AI accounts is actually for business purposes, meaning a significant amount of AI use goes completely unnoticed by businesses.

At the same time, corporate accounts are used for personal matters, which means employees and AI come together wherever convenient, regardless of security policies. In fact, almost half (45.6%) of all personal AI activities take place under licensed packages paid for by companies.

In reality, workers don’t treat business AI and personal AI as separate things, but rather hand off their tasks to whatever AI tool is already open or easily accessible on their device, whether employer-provided or personal, free or paid.

Undetected Personal AI Work Creates Lack of Visibility

Harmonic’s research highlights the visibility gap that is emerging as AI adoption spreads, with legal and governance workers having both the highest usage and highest visibility. These workers account for approximately one-fifth (19.5%) of all AI hours across teams, and 81% of this use occurs on approved tools.

Marketing teams are the second largest users, at 17.5%, but only 39% of GTM AI activity occurs on company-approved tools, leading to poor visibility. But that’s still twice as much visibility as operations teams, where not even a fifth (18%) of activities execute on business plans.

As for why AI is used at work, the clearest goal is efficiency and automation (47%), well ahead of decision support (20%) and risk and compliance (20%). Revenue and growth (7%) as well as innovation (6%) are less common.

The true measurement of usage is minutes, not queries

Where Harmonic’s research differs from other studies is in its use of “minutes” rather than “total queries,” which it says offer a much truer reflection of usage patterns. Longer sessions indicate more data exposure, he says, and Claude comes out on top in actual minutes (10:12) compared to ChatGPT (5:53).

This is particularly problematic when workers choose to use their own personal AI accounts, because sensitive company information and business context remains in their personal AI history even when they leave the company. Organizations don’t even have the legal or technical powers to erase or recover this data, leading to a permanent loss of intellectual property.

The path of least resistance

Harmonic explained that many companies implement strict and clunky authentication processes for enterprise AI tools, making personal tools much easier to use. Popular personal tools like ChatGPT, Gemini, Claude, and Perplexity also require a little more than a Google account (or similar) to log in.

All this while businesses pay extra for licenses that are barely used – Microsoft 365 Copilot is typically deployed at $30 per user per month; ChatGPT Business plans cost between $20 and $25 per month.

“Every organization is investing money in AI right now, and almost none of them know what their employees are actually doing with it,” summarized Harmonic Security CEO Alastair Paterson, noting that this is the first study of its kind to uncover how AI is “actually used at work.”

Clearly, the problem is not necessarily with providing the wrong tools, but rather with ease of access. Going forward, businesses are advised to adopt Universal Single Sign-On (SSO) to make logging in easier. But Harmonic still challenges the “one size fits all” approach, urging employers to consider workflows and provide the right tools to the right teams.

Google logo on black background next to text