Nothing says "the future of work" like turning your job into a leaderboard where forgetting to log your AI usage costs you a career boost.
The Summary
- KPMG rolled out a dashboard in its US advisory division that tracks how often each of its 10,000 consultants use AI, comparing their usage against targets and peer groups
- The firm claims 90% of US employees now use AI weekly, with the dashboard designed to push "frequent and sophisticated" adoption
- Employees report the system is easy to game and creates pressure to hit metrics rather than use AI meaningfully
- This launched in the same division where KPMG recently cut 400 jobs, raising questions about what "AI adoption" really measures
The Signal
KPMG just made the quiet part loud. The Big Four firm installed what amounts to a FitBit for AI usage, letting consultants see exactly how their ChatGPT habit stacks up against their colleagues. The dashboard went live late last year in the advisory division, the same group that just absorbed 400 layoffs.
The official line is predictable: KPMG says regular AI users produce higher-quality work, feel less stressed, and spend more time on strategic thinking. Spokesperson Russ Grote frames it as career accelerant. Use more AI, get promoted faster, serve clients better during their "transformation programs." Classic consulting speak for "our clients are also figuring this out and we need to look like we're ahead."
"Our data shows regular AI users produce higher-quality work, feel less stressed, and spend more time on strategic work."
But here's where the story gets interesting. Two KPMG employees who spoke to Business Insider say the dashboard is trivially easy to manipulate. They asked to stay anonymous because they weren't authorized to talk, which tells you something about the internal climate around this tool.
The gamification is the point. You see your usage. You see your target. You see how you rank against your peer group. It's Strava for PowerPoint optimization. And just like Strava, the people who care most about the metrics are often optimizing for the wrong thing.
Key tensions:
- Pressure to hit usage targets vs. pressure to deliver actual client value
- Public narrative (90% weekly usage!) vs. private reality (easy to game)
- Tool for productivity vs. tool for performance review fodder
Think about what this dashboard actually measures. Not quality of output. Not whether AI made the work faster or better. Just frequency. How many times you opened the tool. Which means consultants now have an incentive to manufacture AI interactions the same way they used to manufacture billable hours.
The Implication
This is the preview of how most big companies will try to manage the AI transition. They'll measure inputs (usage frequency, number of prompts, logins per week) because outputs (better work, faster delivery, smarter strategy) are harder to quantify. And when you measure the easy thing instead of the right thing, you get exactly what you're tracking: high usage numbers and a workforce that's learned to game the system.
If you're building tools for the enterprise AI market, watch this closely. The companies buying your product will want dashboards. They'll want adoption metrics. They'll want proof that the investment is working. KPMG just showed you what happens when you give them exactly that: a tool that tracks everything except whether anyone's actually getting smarter.