Meta built a leaderboard. It ranked all 85,000 employees by how many AI tokens they consumed. Badges included "Token Legend" and "Session Immortal." The top-ranked user burned through 281 billion tokens in a single month — a spend plausibly in the hundreds of thousands of dollars for one person.

The company pulled the leaderboard within days.

Amazon built one too. Called Kirorank, it tracked engineers by AI usage volume. Employees responded predictably: they invented menial tasks for the AI to run, purely to climb the rankings. Amazon shut it down in May after "tokenmaxxing" — the practice of maximizing token consumption as a vanity metric — made the whole exercise meaningless.

"Please don't use AI just for the sake of using AI," Dave Treadwell, an Amazon senior vice-president, told staff in an internal memo reported by the Financial Times.

It's a sentence that captures one of 2026's most expensive management failures.

Uber burned through its entire 2026 AI coding budget by April — four months into the fiscal year. Individual engineers were running $500 to $2,000 per month in token costs. The company capped every engineer at $1,500 per month per tool and started asking where the money went. Uber's COO, Andrew Macdonald, admitted publicly that he could not connect rising Claude Code spending to any measurable business outcome.

AT&T limited employee access to GitHub Copilot. Walmart capped usage of its in-house AI agent. Microsoft canceled most of its Claude Code licenses six months after rolling them out.

The pattern is the same everywhere: companies incentivized volume, got volume, and then couldn't explain what they'd bought.

Glean's 2026 Work AI Index, a survey of 6,000 workers co-authored with researchers at Stanford and UC Berkeley, puts the problem in sharper focus. Workers save roughly 11 hours per week through AI automation. But they spend 6.4 of those hours on what the researchers call "botsitting" — feeding AI tools missing context, checking outputs, debugging mistakes, rerunning prompts, and cleaning up confident-but-wrong answers.

That's 37% of all AI time spent supervising the tools rather than producing with them. Net productivity gain: about 4.6 hours. And 69% of workers admitted to "botshitting" — shipping AI-generated work they hadn't verified and couldn't explain if questioned.

Boston Consulting Group's 2026 Global AI at Work report, which surveyed nearly 12,000 frontline employees, found a parallel problem: 42% of workers reported saving about eight hours per week with AI, but 66% said they received little to no guidance on how to invest the time they saved. Half weren't spending that saved time on anything more strategic.

"Senior leaders are really struggling to articulate what the vision and strategy is on AI," said David Martin, BCG's global leader of People & Organization. "It increases employee fear. It makes it harder for them to understand what objectives they're pushing for."

The mistake wasn't giving employees AI. It was measuring input instead of output.

Token leaderboards are the AI equivalent of measuring programmers by lines of code written. It incentivizes the wrong behavior, rewards waste, and tells you nothing about whether the work improved. Cognition CEO Scott Wu, whose company builds AI coding agents, put it plainly: companies got "carried away" with token tracking and should measure employees on what they produce, not what they consume.

The companies that have avoided this trap share one trait. They didn't roll out AI and ask people to use it more. They identified specific tasks, measured specific outcomes, and pointed the tools at problems rather than at employees.

Viktor lives inside Slack and Microsoft Teams. You @mention it in a thread the same way you would ask a colleague. The output — a PDF, a report, a task created in your CRM, an email drafted in Gmail — lands where it should land.

That distinction matters. There's no token leaderboard. No gamification. No pressure to generate activity for its own sake. You describe a task in plain language. The work gets done. The value is in the output, not the consumption.

You get $100 of free credits to begin. No time limit, no commitment. That's enough to do real work and see what Viktor can actually do before you spend a penny. There's also $50 off your first bill. You must use this exact link to receive both benefits.

The companies in the FT story spent millions learning something straightforward. AI works when you aim it at a specific job. It fails when you tell people to use it more and hope for the best.

The best AI tool is the one that disappears into the workflow. No badges required.

Disclosure: Some links in this article are affiliate links. If you choose to get started with Viktor using the links provided, I may receive a commission — at no additional cost to you. I only recommend tools I use and believe in.