AI credit abuse protection
Meter who uses your AI.Not just how much.
Darkmeter ties every login back to one device, then meters its accounts, locations and credit use, so you can pause the heavy users and keep the real ones.
How Darkmeter sees it
One person. Many accounts. One meter.
Free credits attract people who sign up again and again. Darkmeter recognises the device behind every account and keeps one meter for all of them.
One device · dev_7f3a
ava.k92@mail.test
Pune
ava.k93@mail.test
Frankfurt · VPN
free.7@temp.test
Austin · VPN
Darkmeter
Metering
Credits used
1 account
A dark user signs up and gets free credits.
Illustrative scene. The person, emails and numbers are made up.
The console
See the person behind every account.
The console groups your accounts by the person behind them, shows how likely each label is, and lets you act on it.
Explore the consoleLikely, not certain
Accounts are grouped as likely true users, likely abusers, or unclear. Darkmeter cannot know for certain, so it shows how sure it is.
Evidence for every label
Open any person to see their accounts, the original one, their credit use, and the evidence behind each link.
Act, and teach it
Block or release a person, or mark them as real or abusive. Your decisions become feedback for the model.
Live activity
- NowNew signup · mia.r@mail.test · Toronto
Identity c7d2
Likely true userConfidence HighCredits used · all accounts
0 of 500 · Allow
Accounts · 1
- mia.r@mail.testOriginalToronto
dev_c7d2 · Chrome on macOS
Why Darkmeter thinks so
- One account on this device
Illustrative people, devices and numbers. Select an identity to see its evidence.
Give credits to customers. Not to loops.
We're building Darkmeter with a small group of AI teams. Come shape it with us.