People love AI. ChatGPT has 900 million weekly active users, and 55% of people say they use AI once a week. But despite all that consumer adoption, most people still aren't using AI for meaningful work.
Marc Zao-Sanders has spent two years analyzing tens of thousands of Reddit posts to understand how people actually use AI. He says they’re using it to navigate a messy divorce, to talk to their teenage kid, to meal plan, and to rewrite angry texts to their in-laws.
Our own data shows similar patterns: Most knowledge workers (86%) treat AI as "better Google" for simple lookups. 26% don’t use AI for work-related use cases at all.
This makes most business leaders panic. They say, “Why are we paying for enterprise licenses so people can plan their vacation?”
Marc says, “Slow down.” Personal use cases might not actually be a problem for enterprise ROI. Here’s why.
Can personal AI use cases generate ROI?
Marc still hears very strong reactions to the main finding of his 2025 research – that the #1 AI use case is “therapy and companionship.”
Some people love that use case - others despise it. But mostly, people can’t stop talking about it. And that’s because it embodies the enterprise’s biggest problem: people don’t have nearly as many professional AI use cases as they do personal ones.
“The reaction said more about our discomfort than about the data,” Marc says.
Besides therapy, Marc’s research found that people are using AI for:
- Organizing their lives (calendars, meal planning, household logistics)
- Finding purpose (career direction, life decisions)
- Adjusting email tone (taking the edge off frustrated messages)
- Practicing difficult conversations (rehearsing tough feedback)
- Understanding their legal and financial documents
On the surface, most of these use cases seem purely personal. But Marc says there’s more overlap than you may think.
"Even though these use cases are personal, 90% are still kind of relevant at work," Marc says.
He hypothesizes that when people are practicing “low-stakes” personal AI use cases, they’re building familiarity with AI - and transferable skills - that will also benefit the business.
Here are four work skills that he believes people are building, when they use ChatGPT to text their husband after a fight.
AI aptitude and confidence
Most people who use ChatGPT for personal tasks aren’t following a structured prompt template or “learning” how to use AI in the traditional sense.
But Marc says they are organically discovering the best ways to work with AI. As they use AI to solve problems, they’re learning to provide context, iterate on a recommendation, and upload documents - all the things that drive meaningful output at work as well.
Marc believes personal wins become organic pipelines for professional use case discovery.
“Personal AI use builds fluency because people experiment freely. That fluency leads to meaningful and confident application at work down the line,” Marc says.
Greater emotional regulation
Emotionally regulated employees make better decisions under pressure, handle client stress without breaking down, and give / receive feedback constructively.
Marc says he sees a few transferable use cases in his research:
Use case #3: Finding purpose. People who understand how to use AI to find purpose can become more personally connected to their role, more rooted in their day-to-day routine, and more personally motivated.
Use case #18: Boosting confidence. Those who build self-assurance through personal coaching have a stronger presence in meetings, are more assertive in negotiations, and are more willing to advocate for ideas.
Less tension
Emotional regulation for yourself becomes conflict navigation with others. High-performing teams require psychological safety and trust – and workplace tension destroys both.
Transferable use cases:
Use case #1: Therapy/companionship: When employees process personal frustrations through 1:1 coaching, they’re more able to work through moments of friction with colleagues without being reactive
Use case #29: Deep and meaningful conversations. Developing deeper empathy for others fosters stronger collaboration across different working styles and viewpoints
Stronger organizational capability
Even personal projects can help employees get good at breaking complex initiatives into manageable pieces, juggling competing constraints, and more systematically setting priorities.
Transferable use cases:
Use case #2: Organizing my life: Employees who can manage daily habits can develop more discipline in strategically managing tasks, more accountable to complex projects
Use case #24: Creating travel itineraries: Generating multi-day plans can transfer to coordination of initiatives with multiple constraints and competing priorities
These use cases don’t bring about direct business value. But they may make your team more AI-native employees, methodical workers, stronger communicators, and better collaborators. The invisible ROI might look like less tension, less avoidable project hang-ups, and less burnout.
“These can lead to a more harmonious workplace, but it’s virtually impossible to measure the ROI of that,” Marc says.
Scaling the impact
Using AI for personal reasons allows employees to play, building muscle without any fear of using the tool incorrectly. Failing privately leads them to succeeding publicly at work. In that low-stakes environment, they become good at AI.
Your charter as a leader shouldn’t be to crack down on personal use of AI, but to leverage the satisfaction of personal wins as a way to encourage business exploration.
A few ways you can create that momentum:
- Lead by example: Marc emphasizes the value of storytelling in change management. If employees hear that you are also finding valuable personal use cases with AI, you’re creating psychological safety around this kind of experimentation.
- Encourage employees to share personal AI wins: Marc calls this the ‘jealousy heuristic’. We hear about somebody’s personal success with a new technology and we want it for ourselves. So allow personal use cases to be shared during AI shout-outs so other employees can practice the same skills and experience the same wins.
- Seed the conversation with inspiration points: As part of your own narrative, demonstrate how your own personal AI use case applied to a valuable team workflow. It’s as simple as “I tried AI for XYZ and it got me thinking this would make a lot of sense in ABC workflow…”
“If people can see that an AI use case really affected your life in a believable, tangible, concrete, way, they’re going to find a way to do that for themselves,” Marc says.
TL;DR – Marc says don’t be a token pincher. All AI experimentation has value to you at this point. It’s up to you to shape it into something more impactful.
Interested in Marc's AI in the Wild research? Reach out to him directly.
