Skip to main content
SecureAI Guard

Resources

Plain explanations of the attack classes, controls and standards that apply to LLM applications. No gated PDFs.

  • Agent security and excessive agency

    An agent turns text generation into action. What excessive agency means, why shared service accounts are the core defect, and the controls that hold.

    · 4 min read
    agent security
    excessive agency
    OWASP LLM Top 10

  • AI red teaming for LLM applications

    How to red team an LLM application rather than a model: scope it to your system, build the test set from ATLAS and OWASP, and measure something durable.

    · 4 min read
    AI red teaming
    MITRE ATLAS
    testing

  • How LLM applications leak sensitive data

    The paths by which data actually escapes an LLM application: provider logs, retrieval, conversation state, agent tools, telemetry, fine-tuning and the index.

    · 4 min read
    data leakage
    OWASP LLM Top 10
    data privacy

  • LLM guardrails, and what they can and cannot enforce

    What an LLM guardrail is, the four places one can run, why a model guarding a model shares the same weakness, and how to tell a control from a suggestion.

    · 4 min read
    LLM guardrails
    prompt injection
    application security

  • PII redaction for LLM applications

    Keeping personal data out of a third-party model without breaking the product: detection limits, redaction versus tokenisation, and when it is wrong.

    · 4 min read
    PII redaction
    data privacy
    GDPR

  • RAG security: how retrieval leaks data

    Retrieval moves the security question from the model to the corpus: an index with no access control, poisoned documents, and a second copy of your data.

    · 4 min read
    RAG security
    vector database
    prompt injection

  • The OWASP LLM Top 10 (2025), entry by entry

    All ten entries of the OWASP Top 10 for LLM Applications 2025, what each one actually means in a deployed system, and which control closes it.

    · 4 min read
    OWASP LLM Top 10
    application security
    risk

  • What is prompt injection?

    Prompt injection is what happens when untrusted text reaches a model's instruction context. Direct and indirect forms, why filtering fails, and what helps.

    · 4 min read
    prompt injection
    OWASP LLM Top 10
    application security