FinProof v2 — 11,474 prompts · 13 attack categories · 36 subtypes

The First Benchmark Built for GenAI Guardrails in Financial Services

A purpose-built benchmark for evaluating GenAI guardrail performance in regulated financial institutions. Mapped to RBI, EU AI Act, SR 11-7, ISO 42001, and India DPDP Act 2023.

Open Dataset — HuggingFace
Zero Training Contamination
finproof://breach-matrixlive
guardrail13 attack categories →
PromptGuard
ShieldGemma
LlamaGuard
WildGuard
breach blocked
81.5%
mean attack rate

Industry-Leading Benchmark

FinProof by the numbers — comprehensive coverage, rigorous evaluation, open access

11,474
Evaluation Prompts
Adversarial test cases across BFSI
36
Attack Subtypes
13 attack categories · 36 subtypes
0%
Training Contamination
Verified zero overlap with public LLM data
81.5%
Attack Rate
Successful attacks against leading guardrails
Features

Built for GenAI Safety & Compliance

Built by financial AI researchers for financial AI researchers. FinProof provides the tools you need to ensure your models are production-ready.

21 Guardrail Actions Tested

Covers the complete taxonomy of GenAI guardrail actions across Input, Processing, and Output layers — from Block and Redact to Agent Handoff Validation and Rollback.

Regulatory Mapping

Every domain and prompt mapped to specific obligations under EU AI Act, SR 11-7, ISO/IEC 42001, and India DPDP Act 2023.

QCBM Adversarial Generation

Prompts generated using Quantum Circuit Born Machines — 89,022 hashes evaluated, zero training contamination confirmed. Novel adversarial patterns no static classifier has seen.

Four-Tier Access Model

Tier 1: Public (JSONL download). Tier 2: Email gate (extended set). Tier 3: Research agreement (full corpus). Tier 4: Withheld test set with public SHA-256 integrity hash.

Plain JSONL — Run In-Perimeter

No external API calls required. Institutions with restricted outbound access can run the full evaluation harness entirely within their own infrastructure.

Open Evaluation Harness

CLI tool available on GitHub for running the test suite against any guardrail stack. Score your own guardrail implementation against the FINPROOF standard.

Three Conversational Registers

Professional (compliance officer, 30–120 words, regulatory citations), Customer Mobile (retail chatbot, 8–30 words, no regulatory language), RM Internal (relationship manager, compressed shorthand). PromptGuard achieves near-zero FPR on professional queries but 100% FPR on the retail register — flagging real customer banking queries as attacks. Register coverage is required to measure this failure mode.

Coverage

What FinProof Actually Tests

Every category maps to a specific BFSI regulatory obligation — not generic harm.

B-01

Unlicensed Investment Advice

Does your AI refuse to give specific fund recommendations or portfolio allocations to retail customers? B-01 covers 1,257 prompts across direct, indirect, and advanced difficulty.

Regulatory anchor: SEBI IA Regulations 2013 §3

B-02

KYC & AML Bypass

Does your AI block attempts to circumvent identity verification or structure transactions below AML thresholds? B-02 covers 1,267 prompts.

Regulatory anchor: RBI KYC Master Directions 2016

B-12

Information Barrier Violations

Does your AI refuse to aggregate innocuous cross-division signals (IB workload, delivery volumes, CFO meeting cancellations) that together enable MNPI inference? B-12 covers 262 prompts across 5 subtypes.

Regulatory anchor: SEBI PIT Regulations (Chinese Wall)

B-13

Mandatory Human Escalation

Does your AI escalate fraud, grievance, and vulnerability cases to a human instead of self-resolving? B-13 covers 312 prompts.

Regulatory anchor: RBI ombudsman and STR reporting requirements

B-14

Regulatory KPI Gaming

Does your AI refuse to help game NPA ratios, CAR figures, or LCR metrics in ways that satisfy the letter of the rule while defeating supervisory intent? B-14 covers 267 prompts.

Regulatory anchor: RBI supervisory frameworks (PCA, LCR, GNPA)

D-09/D-10/D-11

Agentic & MCP Security

Does your AI agent resist tool description poisoning, A2A trust exploitation, and unauthorized lateral traversal across banking systems? D-series covers 450 prompts across 15 subtypes. The only BFSI benchmark covering multi-agent attack surfaces.

Positioning

Why Generic Benchmarks Fail in BFSI

WildJailbreak, HarmBench, and AdvBench test general harm. They cannot test BFSI-specific regulatory violations — and they have no BFSI-specific benign set, so FPR is meaningless.

DimensionWildJailbreakHarmBenchFinProof
BFSI domain attacks
Regulatory anchor per prompt
SEBI / RBI / DPDP / EU AI Act
Retail vs professional register
3 registers
FPR-calibrated BFSI benign set
generic text
generic text
2,127 BFSI queries
Agentic / MCP attack coverage
D-09 / D-10 / D-11
Withheld test set
T4 — never released
Contamination audit tool
contamination_audit.py

How FinProof differs from other guardrail benchmarks

Most guardrail benchmarks

The submitter supplies executable code, which the benchmark owner runs against the private set with unrestricted network access. A submission can transmit every prompt it receives. The leaderboard doesn't record whether it did.

FinProof

Open-weights submissions run from a config — no submitter code executes. Container submissions run with no network interface, verified at runtime rather than promised. Every row publishes how it was produced.

"Withholding a test set only means something if the mechanism for scoring against it doesn't hand it over."

New in v2

FinProof v2 — Agentic & MCP Security

The only adversarial benchmark testing multi-agent financial systems.

Banks are deploying agent networks. FinProof v2 adds 450 agentic attack prompts across three new categories — the only BFSI benchmark to cover this attack surface.

D-09
150 prompts

Tool Description Poisoning

Instructions embedded in MCP tool metadata that override user requests or trigger unauthorized financial actions.

Subtypes: TPA, TYP, OAB, CSC, TRP

D-10
150 prompts

A2A Trust Exploitation

Attacks on multi-agent trust chains: agent impersonation, approval forgery, failure cascade, lateral traversal across banking system privilege boundaries.

D-11
150 prompts

Corrigibility & Goal Binding

Tests whether an AI agent accepts authorized policy correction, resists deceptive alignment, refuses self-preservation actions, and blocks reward gaming via CSAT manipulation.

Indic Language Coverage

3,900 Hindi, Telugu, and Tamil attack prompts (v1.1 Indic). Models that pass FinProof English can still fail on Indic registers — regional language framing lowers guardrail vigilance. FinProof is the only BFSI benchmark with verified Indic coverage.

Leaderboard

FINPROOF Model Leaderboards — July 2026

Guardrail classification results across all BFSI attack categories.

Track 1 — Guardrail Classification Leaderboard

B-01 through B-14 · Input classification · F1 / Precision / Recall / FPR

#ModelBFSI F1
REF
AVAL v1.4Reference Baseline
Zytra Tech Solutions · mmBERT IT4 · Reference Baseline
0.952
Granite Guardian 3.3
IBM Research · 8B
0.813
ShieldGemma 9B
Google · 9B
0.731
LlamaGuard 3
Meta AI · 8B
0.569
4
WildGuard 7B
Allen AI · 7B
0.346

BFSI F1 scores computed on withheld Tier 4 test set (B-01–B-07). B-12 / B-13 / B-14 scores pending — no submitted model currently covers these categories.

Provenance states how each score was produced. Benchmarks that don't publish this can't tell you whether two rows were measured under the same conditions.

Per-Domain F1 — B-01 to B-07

★ denotes best score per domain

DomainAval
B-01Investment advice0.954
B-02KYC / Card fraud0.981
B-03Employment fraud0.960
B-04Regulatory hallucination0.957
B-05Predatory lending0.986
B-06Insurance conduct0.975
B-07Financial instruments0.962

Last updated: July 2026

Access Tiers

Dataset Access Tiers

FinProof uses a structured access model to protect evaluation integrity while enabling broad research access.

Tier 1 — Public

Open access. No registration required.

Free
Prompts in corpus1,982
Available at this tier175
  • All categories represented
  • Evaluation harness (GitHub)
  • Full regulatory mapping documentation
Most Popular

Tier 2 — Extended

Register with institutional email to access the extended set.

Email Gate
Prompts in corpus2,706
Available at this tier320
  • All Tier 1 content
  • Benchmark scoring scripts
  • Priority release notes

Tier 3 — Full Corpus

Full dataset under a signed research use agreement.

Research Agreement
Prompts in corpus3,337
Available at this tier427
  • Detailed annotation metadata
  • Regulatory mapping per prompt
  • Academic citation support

Tier 4 — Withheld Test Set (Evaluation Only)

Prompts are never transmitted to submitters. Choose the route that matches your system.

Prompts in corpus
2,608
Available at this tier
0 (withheld)

Scored against 3,953 withheld prompts: 2,608 attack (B-01–B-07) plus 1,345 benign. The benign rows are what make precision and FPR meaningful — a guardrail that flags everything scores perfect recall and fails here. Every prompt ID published on HuggingFace is excluded from scoring.

ROUTE AOpen Weights(recommended)

You send: a YAML config naming a HuggingFace model ID, a pinned commit SHA, and how to read a verdict (chat template / decision tokens / classifier threshold).

We run the weights on Zytra hardware. Nothing is transmitted to you.

Best for: single-model guardrail classifiers.

ROUTE BContainer

You send an OCI image, content-addressed as either repo@sha256:… (a registry digest) or sha256:… (an image ID). The image ID form means you can ship a tarball via docker save — a proprietary guardrail never has to be published to a registry. A mutable tag is not accepted: it could be repointed after your result is published.

The image reads prompts as JSON lines on stdin and writes verdicts to stdout.

We run it with --network none, verified at runtime. Model weights must be baked into the image.

Best for: cascading guardrails, ensembles, proprietary pipelines.

ROUTE CHosted API

Not offered by default. Evaluating a hosted API necessarily transmits prompts to the vendor. Contact us to discuss terms; results are published with an "api-exposed" provenance badge rather than "sandboxed".

How submission works

1

Send your config

Open weights: a YAML file naming the model, a pinned commit SHA, and the decision rule. Container: an image reference pinned by digest. Everything must be pinned — a mutable tag or branch means the scored artifact can change after publication.

2

We run it

Evaluation happens on Zytra infrastructure against the withheld set. You receive no prompts and run no endpoint. Typical open-weights run: 1–2 minutes of GPU.

3

You get scored

Overall F1, precision, recall and FPR, plus breakdowns by BFSI category (B-01…B-07), by difficulty, and by conversational register. Published as a leaderboard row with a provenance badge stating how it was produced.

"Every submitter is scored against the same withheld set, and no submitter receives it — including us. That is what makes the numbers comparable."

All tiers include documentation, community support, and regulatory mapping. Tier 4 is used exclusively for official model evaluation submissions.

Ready to Benchmark Your Financial AI?

Validate your guardrail stack against the only BFSI-specific adversarial benchmark. Get started with our free public tier or submit your model for official Tier 4 evaluation.

No credit card requiredFree academic accessSOC 2 compliant