Use case · AI credit farming
Stop bleeding GPU dollars to credit farms.
Every AI tool fights the same trade: $0.10 rents a throwaway cloud server that farms an account worth $5–$30 in resold AI usage. Root Herald raises the cost of an account to the cost of a real, physical machine, and the resale stops paying off.
Economics
What a fake account earns vs. what it now costs.
A fake account is worth the credits a farmer can drain and resell: $5–$30 on typical platforms. With Root Herald, making one costs a real $30+ machine. The crossover starts at the low end of that range and only widens above it. Farming stops turning a profit.
AI free-tier credit farming
Suno, Lovable, Krea, Civitai
Per-identity yield
$5 – $30
Rational ceiling
< $10
Only Tier 1 was profitable; Tier 3+ exceeds yield.
What stops working
The cheap paths, closed under strict policy.
Rented cloud servers
A cloud server dressed up as a PC can't prove it's real hardware. Amazon, Azure, and Google virtual machines are all turned away by the default policy.
Software emulators
A 'chip' that's really just software in a container can't prove it's a physical part. Rejected. It even labels itself as a fake.
Anti-detect browsers and proxies
Tools that fake a fresh browser fingerprint don't create a fresh chip. One machine keeps one anonymous device ID, no matter how many disguises it wears.
Throwaway emails and phones
Burner emails and SIM-farm numbers don't create new hardware. The device ID stays the same across every rotation.
Integration
Check at signup, then at each credit grant.
Check the device at signup to gate account creation, then check again each time you hand out free credits, so account takeover or device-sharing surfaces. The anonymous device ID is stable per customer; bind it to the user record.
import { RootHerald } from "@rootherald/node";
const rh = new RootHerald({ secretKey: process.env.RH_SECRET_KEY }); // rh_sk_…
// Earlier: rh.issueChallenge() -> send the one-time challenge code to your app,
// which collects the sealed hardware proof and posts back { challengeId, evidence }.
const verdict = await rh.verify(evidence, {
challengeId,
policy: "rootherald:builtin:strict-hardware", // our "real physical chips only" rule
});
if (verdict.device.verdict !== "pass")
return res.status(403).json({ error: 'device_check_failed' });
const existing = await db.users.findByDevice(verdict.device.ueid);
if (existing) return res.status(409).json({ error: 'device_already_registered' });
await db.users.create({ deviceId: verdict.device.ueid, email });Real-world precedents
The industry signal.
Major AI providers have pulled back automatic free credits and banned third-party tools, citing credit farming that makes free promotions unviable. Arkose Labssells a product aimed squarely at "LLM platform abuse"; the category exists because the problem is that widespread. Civitai, Lovable, Suno, Krea, and Janitor AI all sit squarely in the credit-farming crosshairs.
Keep the free tier. Lose the abuse.
Free up to 10K device checks a month. Check the device at signup, and farming a free account costs a real machine.