Prompt Injection Defence in Depth 2026: Six-Layer Architecture for LLM Security

Q: What does Layer 1 (input sanitisation and normalisation) actually deliver, and what are its known limitations?

Layer 1 normalises the input into a canonical form and strips the channel-level vectors. Three engineering deliverables. Unicode and encoding normalisation: the input is normalised to NFC or NFKC depending on the application context; the homoglyph attacks (Cyrillic substituted for Latin to evade regex filters, mathematical alphanumeric symbols substituted for ASCII letters), the bidirectional-control character attacks (right-to-left override embedded to obscure what the user is approving), and the zero-width character attacks (joiners and non-joiners inserted to break pattern matching) are addressed at this layer. Length and structural validation: inputs above the application's defined maximum are rejected before consuming LLM context; structurally-typed inputs (JSON against schema, YAML parsed and re-emitted in canonical form, query inputs against the application's permitted grammar) are validated, removing the smuggling vectors that depend on parser-disagreement between the validation layer and the downstream LLM. Known-pattern stripping for the injection-pattern catalogue: published collections of jailbreak prompts, the application's own incident-derived patterns, threat-intelligence-fed patterns are matched and either stripped, escaped, or rejected. The layer's known limitations are deliberate — catalogue-based filtering does not catch novel patterns, and the architectural intent is for Layer 1 to be the easy-to-evade frontline whose role is to catch trivial attacks and keep the more expensive layers below their attack volume. The engineering investment is modest; the layer's value is the consistency of the normalised input the layers below see. Teams that skip Layer 1 ship the application that produces inconsistent behaviour as the input format varies, which is a quality issue regardless of the security framing.

Satyam Kumar

返回博客

ai-architecture

Prompt Injection Defence in Depth (2026): Six Layers from Input Sanitisation to Output Firewall

By Satyam KumarMay 13, 202622 min read

prompt injection ai security defence in depth llm security output firewall tool use authorisation capability matrix spotlighting injection detector secondary review llm egress filter action effect simulator rag security agent security jailbreak defence guardrails ai architecture 2026

Prompt Injection Defence in Depth (2026): Six Layers from Input Sanitisation to Output Firewall

Frequently Asked Questions

分享这篇文章

Twitter LinkedIn WhatsApp

Satyam Kumar

Founder & AI Architect, AppScale LLP

人工智能和云架构师。帮助团队构建可扩展到数百万的系统。

LinkedIn GitHub

Prompt Injection Defence in Depth (2026): Six Layers from Input Sanitisation to Output Firewall

Frequently Asked Questions

分享这篇文章

Comments

Leave a comment