What is an LLM vulnerability scanner?

An LLM vulnerability scanner sends batteries of adversarial probes — jailbreaks, prompt injections, PII extraction attempts, harmful-content requests — at a target LLM and grades the responses. The output is a vulnerability + optimization report with per-probe evidence, severity, and a prioritized fix list.

Which LLMs can I scan with FilterPrompt?

OpenAI, Anthropic, Google Gemini, Azure OpenAI, plus any OpenAI-compatible endpoint — Ollama, Groq, Mistral, Together AI, OpenRouter, Perplexity, Hugging Face, vLLM, or your own custom endpoint. Bring your own keys per tenant.

What kinds of vulnerabilities does FilterPrompt test for?

Jailbreaks (DAN, role hijack, translation smuggling), direct and indirect prompt injection, system-prompt extraction, harmful-content compliance, PII / secret leakage, bias & fairness, RAG poisoning, agent/tool abuse, output quality, and robustness — categories map to the OWASP LLM Top 10.

How are probes graded?

Each probe declares an evaluator: regex match, refusal-check, contains-check, or an AI judge (Gemini 3 Flash). Pass/fail comes with severity, category, the exact prompt sent, the model's full response, and the evaluator's reason — fully auditable.

How much does a scan cost?

1 credit per probe executed. New accounts get 1 welcome credit on signup. Pay-as-you-go credit packs after that — credits never expire. Connecting LLMs and creating tenants is free.

HIPAA, GLBA, SEC and PCI DSS: The Sector Rules That Already Force AI Scanning in Healthcare and Finance

Regulation · 2024-12-03 · 11 min read · FilterPrompt Compliance Team

There is no 'AI law' for U.S. healthcare or finance — but HIPAA, GLBA, SEC cyber-disclosure and PCI DSS already treat a leaky chatbot as a breach. Here's how to test for it.

Outside horizontal AI laws, sector regulators already have binding data-protection rules that apply the moment an organization connects sensitive data to an LLM. These rules don't mention 'AI' by name, but a chatbot that leaks a medical record or a customer's account number is a violation under existing law, regardless of whether an AI model was involved. Compliance teams don't get to wait for an AI-specific statute.

The sector rule map

Healthcare: HIPAA in an LLM world

The HIPAA Security Rule requires covered entities and business associates to implement administrative, physical and technical safeguards for electronic PHI. When you feed a hospital's ticketing data into a summarization LLM, three things become in-scope: whether PHI reached an unauthorized model, whether the model can be jailbroken into revealing PHI it saw, and whether audit logs would let you prove either way. All three are LLM vulnerability categories — sensitive data exposure (LLM02), system prompt leakage (LLM07) and prompt injection (LLM01).

Financial services: GLBA, SEC and the 'material incident' clock

The FTC's amended Safeguards Rule (in force since June 2023) requires financial institutions to design, implement and maintain a written information-security program with continuous monitoring or periodic penetration testing and vulnerability assessments. A customer-support LLM is exactly the kind of system this covers. On top of that, the SEC's 2023 cybersecurity rules require Item 1.05 8-K disclosure within four business days of determining a cyber incident is material. An AI leak of customer account information starts that clock.

Payments: PCI DSS v4.0 and the PAN problem

PCI DSS v4.0 Requirements 3.3 and 3.5 prohibit the storage or transmission of unmasked Primary Account Numbers except where strictly necessary and with strong cryptography. A chatbot that echoes a card number in its response, or a RAG pipeline that indexes an unredacted transaction log, both fail. The mitigation is a pre-prompt DLP layer — redact PAN, CVV, expiry and cardholder name before the request reaches the model, and re-scan the response before it reaches the user.