# FinProof (finproof.ai)

Adversarial test suite of 11,474 prompts that scores how well guardrails on banking AI agents block regulated-finance attacks.

- Atlas card: https://inite.ai/en/atlas/finproof.ai
- Site: https://finproof.ai
- For: Teams in regulated financial institutions evaluating genAI guardrails before production deployment
- Categories: dev
- Use not measured/100 · alive 100/100

## Used for

- Measure guardrail refusals against banking-specific adversarial prompts
- Test agent tool poisoning and multi-agent trust attacks
- Check guardrails don't over-block legitimate retail customer queries

## Price

- no price printed (read 2026-10-29 on https://finproof.ai/#pricing)

## For developers

None found.
- Read 2026-10-29

## Your data

- Certifications stated: none
- Read 2026-10-29

---

Source: INITE Atlas, https://inite.ai/en/atlas/finproof.ai. JSON: https://inite.ai/api/atlas/finproof.ai. MCP: https://inite.ai/api/mcp (atlas_lookup).