Data shows workers are using AI on the sly

Every day might be Take Your AI to Work Day if you’re among the three out of four knowledge workers who use generative AI tech at their job—with or without a boss’s go-ahead.

The stat comes from new research from Microsoft and LinkedIn, which found that an overwhelming number of those AI-savvy workers (78%) are using the tech without the permission or knowledge of their employers. A little over half of these users were reluctant to admit they use AI for important tasks and worry it makes them look replaceable.

This year’s Microsoft Work Trend Index was an effort to understand how chatbots have begun to seep into the workplace. It comes as the tech giant pushes its own AI-powered Copilot software across its productivity suite and as a standalone product. Microsoft isn’t alone; enterprise companies like Salesforce, SAP, and Adobe have introduced all-purpose AI tools that aim to expedite day-to-day office chores.

A pressure valve? But for all the new enterprise AI products gunning for a corporate contract, AI adoption is more often coming from individual employees, potentially because they are struggling under the weight of mounting workloads, according to Microsoft Senior Director of Research Alexia Cambon. Meanwhile, many leaders are struggling to understand how to implement AI and could be lacking a guiding vision, she said.

“We know from prior reports [that] employees are incredibly overwhelmed with the pace and volume of work. We know that the work environment has evolved very quickly in a short span of time and essentially has outpaced employees’ ability to keep up with it,” Cambon told Tech Brew. “They’re really turning to those tools for relief within that very high-pressure work environment.”

However, unsanctioned use could open companies up to security dangers if workers drop proprietary data or information into chatbots. “The high risk with BYOAI is, of course, that you are not operating it within a safe environment; you’re feeding information into tools that aren’t necessarily protective,” Cambon said. “And that is something that we know leaders are very worried about. All of that uncertainty is stalling innovation and implementation.”

Meet the bot buffs: AI use at work has accelerated in recent months, and adoption tends to cut across all generations, according to the report. Around 46% of those who said they use AI at work only started using it less than six months ago. The percentage of workers who said they bring their own AI to work were highest among Gen Z workers (85%), then descended with age (78% of millennials, 76% of Gen X, and 73% of boomers).

The report sorts users into four buckets based on attitudes toward AI and usage patterns: skeptics, novices, explorers, and power users. Power users tend to tap the tech several times a week and reported saving more than 30 minutes a day as a result, according to the report.

Skeptics, on the other hand, may use AI, but generally distrust it, Cambon said. This is the audience that some companies will ultimately look to convert to power users, she said.

“[Skeptics are] actually quite frequent users of AI, but don’t like it,” Cambon said. “What we’re seeing with skeptics is all they need is one bad experience with AI to kind of swear it off for good.”

Cambon said data collected from this report will ultimately help Microsoft’s product team understand how it can improve the Copilot product it offers businesses and consumers.

“If the big takeaway is [that] employees really want AI at work and they’re moving really quickly ahead,” Cambon said. “Our mission [is] to help companies catch up to them and equip them to be able to do it in a safe, sanctioned way.”