Post B2JtR3pyGJUXoXeWZs by mttaggart@infosec.exchange
(DIR) More posts by mttaggart@infosec.exchange
(DIR) Post #B2JtR3pyGJUXoXeWZs by mttaggart@infosec.exchange
2026-01-15T05:24:38Z
0 likes, 1 repeats
Problem: LLMs can't defend against prompt injection.Solution: A specialized filtering model that detects prompt injections.Problem: That too is susceptible to bypass and prompt injection.Solution: We reduce the set of acceptable instructions to a more predictable space and filter out anything that doesn't match.Problem: If you over-specialize, the LLM won't understand the instructions.Solution: We define a domain-specific language in the system prompt, with all allowable commands and parameters. Anything else is ignored.Problem: We just reinvented the CLI.
(DIR) Post #B2JtR4zZy4ltOd9iC0 by wolf480pl@mstdn.io
2026-01-15T20:17:52Z
0 likes, 0 repeats
@mttaggart sql injection dejavu.I wonder if PSTN also went through a similar cycle before they discovered out-of-band signaling.
(DIR) Post #B2JtR9cMloArk5KkCm by mttaggart@infosec.exchange
2026-01-15T05:33:54Z
0 likes, 0 repeats
What are we doing with our time on this earthhttps://www.promptarmor.com/resources/claude-cowork-exfiltrates-fileshttps://www.varonis.com/blog/reprompt