Ground truth on AI-written text.

Truth in AI, by not using AI. On the public RAID benchmark, we beat every published detector on adversarial paraphrase — across three typologically distinct languages, with results anyone can verify in one command.

Get API key — first 10 free See it verify a result
0.9302
Mean AUC
across 30 attack × domain configs on RAID.
1.0000
German homoglyph AUC
CI [1.0000–1.0000]. Perfect separation.
99.2%
TPR at 5% FPR
Paraphrase attack. The operating point institutional buyers actually use.
0.0016
Expected Calibration Error
Confidence scores you can defend in policy decisions.

Not a black-box guess.

Every result is Ed25519-signed and independently reproducible on your machine. Anyone can verify without your permission.

Not a language model.

CPU-only. Under 10 ms per document. No LLM in the loop, no per-query API cost, no hallucination surface in the detector itself.

Language-agnostic by design.

Validated on Germanic, Slavic, and code — not just English. EU AI Act deployers and multilingual institutions no longer have to choose.

Built for institutional use.

Purchase orders welcome. SOC 2 posture documented. FERPA-aligned data handling. On-prem or air-gapped — the detector runs anywhere Python does.

The full RAID breakdown.

AUC on the three languages we currently ship, against each of the RAID adversarial attacks. The paraphrase attack is where every competing detector collapses to 0.40–0.60 AUC. Homoglyph is the classic Unicode-obfuscation adversarial threat model.

Language / domain Baseline (none) Paraphrase Homoglyph Synonym
Code 0.879 0.971 0.9999 —
German 0.916 0.9958 1.0000 0.927
Czech 0.943 0.998 0.9958 —

Every cell is an area under the ROC curve on the RAID benchmark (Dugan et al., ACL 2024). Full 30-config table, confidence intervals, TPR‑at‑FPR curves, and ECE calibration numbers are in the signed benchmark result — verifiable in one command below.

Don't take our word. Verify.

The benchmark result is Ed25519-signed, its dataset manifest is SHA-256 checked against the public RAID release, and the artefact itself is Bitcoin-anchored via OpenTimestamps. Anyone can reproduce every AUC on this page — without our permission, without an NDA, without an API key — in one command in a terminal: 

# clone the signed benchmark, verify signature + dataset + AUCs
git clone https://github.com/trigeochiral/truth-in-ai-proof
cd truth-in-ai-proof
pip install truth-in-ai-verify
truth-in-ai-verify --result RAID.signed.json --pubkey pubkey.pem
# → ✓ RESULT VERIFIED  ·  Mean AUC 0.9302 across 30 configs
Open the proof repo →

$0.05 per verified result.

First 10 calls free. No credit card. Verify our claim on your own texts before you buy.

Get API key

Institutional purchase orders, OEM licensing, and air-gapped on-prem deployments welcome — [email protected]

Talk to David.

20-minute call. Integration, purchase orders, custom terms, cleared-work availability.

Email David to book →

Or reach us directly — [email protected]