Engineer fired for low AI token usage despite code quality
Is this a scandal?
No longer — the story has resolved. Noise 4/100, cooling down, across 0 sources.
Enterprises will likely abandon raw token volume KPIs in favor of outcome-based AI metrics because current incentives demonstrably reward waste and sabotage over genuine productivity gains.
Noise 4/100 — louder than 96% of tracked AI controversies.
Why it matters
Tying employment to raw token volume incentivizes wasteful AI abuse over productivity, signaling a dangerous misalignment between management KPIs and actual software engineering value.
Key points
- Engineer terminated after three months at bottom of internal AI token usage leaderboard despite active tool adoption.
- Coworkers allegedly inflated metrics by inducing model thinking loops, spending hundreds of thousands monthly versus tens of thousands.
- Internal tracking measured total tokens consumed rather than output quality or business value delivered.
- Employee claims sabotage occurred when a peer edited system prompts to block high-token reasoning strategies.
- Leadership reportedly mandated terminations for low AI usage to enforce organizational AI adoption targets.
- Incident illustrates risks of using raw consumption volume as a primary proxy for engineering productivity.
The story
A software engineer reported being terminated after ranking last on an internal AI usage leaderboard for three consecutive months, despite attempting to increase token consumption through verbose coding and extended model reasoning. The employee alleged that colleagues artificially inflated usage by trapping models in thinking loops, generating hundreds of thousands of dollars in monthly costs compared to the engineer’s tens of thousands. Management reportedly enforced these metrics to identify the most AI-adapted teams, leading to the dismissal of low-volume users regardless of output quality. The engineer further claimed a coworker sabotaged their system prompts to prevent high-token reasoning strategies. This incident highlights emerging labor disputes where quantitative AI adoption metrics supersede traditional performance evaluations, raising concerns about perverse incentives in enterprise AI integration.
Who's involved
Alleges termination resulted from flawed metrics and coworker sabotage rather than legitimate performance issues.
Reportedly directed terminations of low-AI-usage staff to ensure organization became the most AI-forward unit.
Allegedly executed termination due to binding directives from upper leadership despite potential reservations.
Noise Level
The timeline
Public account posted to Twitter
CathPoaster shares detailed allegations of metric gaming and wrongful termination.
Termination executed after third low rank
Engineer dismissed following skip-level directive to cut lowest AI users.
System prompt sabotage alleged
Coworker reportedly modifies prompt to block high-token reasoning strategies for five days.
First month of low ranking recorded
CathPoaster finishes last despite using AI for all code generation.
AI usage leaderboard tracking begins
Company implements internal metric ranking engineers by total AI tokens consumed.
The full record
Sources & methodology
Every claim above traces to these primary items. How we score →
The forecast
Enterprises will likely abandon raw token volume KPIs in favor of outcome-based AI metrics because current incentives demonstrably reward waste and sabotage over genuine productivity gains.
Forecast, not fact — an editorial estimate we score when this resolves.
That's the complete picture as of — nothing more to know right now. We'll update this page the moment it changes.
Join the Discussion
Discuss this story
Community comments coming in a future update
Be the first to share your perspective. Subscribe to comment.