The UnboundCompute blog
Writing on web security
-
Multimodal Prompt Injection: Hiding Instructions in an Image
Multimodal prompt injection hides instructions inside an image or PDF so a vision model reads them as commands while text filters never…
-
Glitch Tokens: The Words That Break a Language Model
A glitch token attack feeds a language model rare, under trained tokens to break it. Learn how these dead tokens form, why…
-
Markdown Image Exfiltration: How a Chat UI Leaks Your Data
Markdown image exfiltration lets a chat assistant leak private data through an auto rendered image URL, no click needed. See how it…
-
LLM Backdoors: Hiding a Trigger in the Training Data
An llm backdoor attack plants a hidden trigger during training so a model looks aligned but flips on command. See how it…
-
Tool Output Injection: When an Agent’s Own Tools Lie to It
Tool output injection is when an AI agent trusts attacker controlled data from a tool and follows hidden instructions. How it works…
-
Embedding Inversion: Reading Text Back Out of a Vector Database
An embedding inversion attack reads the original text back out of stored vectors. See how a leaked vector database can leak plaintext,…
-
MCP Token Passthrough: How an Agent Hands Over Its Access
MCP token passthrough lets a server reuse a token it was never issued. See how this confused deputy bug works and how…
-
Adversarial Suffix Attacks: The Gibberish String That Jailbreaks a Model
An adversarial suffix attack appends a string of nonsense tokens to flip a model from refusing to complying. See how it works…
-
Prompt Injection to XSS: When Model Output Becomes the Payload
Prompt injection to XSS turns a hidden instruction into script that runs in your users’ browsers. See how the chain works and…
-
CSS Injection: Stealing Data With Style Rules and No JavaScript
CSS injection lets an attacker leak a CSRF token character by character using only style rules, with no JavaScript. See the technique…