Un-Jailbreakable AI Doesn't Exist—But Open, Neural-Symbolic Gets Closest

Perfectly "un-jailbreakable" AI models don't exist—it's an unrealistic goal. But the best way to get close is neural-symbolic AI combined with open-source models, not closed proprietary systems.


The real threat isn't simple prompt injection, but "capability-elicitation attacks"—where an AI follows instructions but is gradually coaxed over hundreds of prompts into producing something dangerous.


The solution: a "generate, then verify" pipeline. Let the neural model generate outputs, but quarantine risky ones and pass them through a symbolic verification layer (formal analyzers, sandboxes, logic engines) that rigorously judges what the output actually does. This is more reliable than just using another LLM to check things.


Why openness helps: An open ecosystem can field a diverse ensemble of specialist verifiers—far better than any single company. Independent verifiers with different blind spots make the system harder to game. While bad actors can still download open models, gating the main public deployment shapes most usage toward legitimate defenders, who then outpace attackers.


Caveat: This works well for defense-symmetric domains like cybersecurity. For offense-dominant, irreversible threats like biowarfare, no architecture makes this easy.


Bottom line: For most practical threats, openness plus neural-symbolic AI is safer than closed, centralized models.

https://bengoertzel.substack.com/p/un-jailbreakable-ai-models-arent

Comments

Post a Comment

Popular posts from this blog

Prompt Engineering Demands Rigorous Evaluation

In the blog post “Prompt Engineering Requires Evaluation” on the Shostack + Associates website, the author argues that treating prompts for large language models (LLMs) merely as creative artefacts is insufficient. Engineering prompts properly demands structured evaluation frameworks — what the AI community calls “evals” — to test which prompt versions work better, with which models, and under which conditions. The post highlights that simply assuming a prompt is “good enough” creates risks when LLMs are integrated into production systems (e.g., for threat modeling). It advocates for measuring prompt performance, variation effects, and tool-chain dependencies (model, context, ancillary materials).  Ultimately the message is: prompt engineering should borrow disciplined practices from software engineering (versioning, testing, benchmarking) rather than relying on informal experimentation.  https://shostack.org/blog/prompt-enignieering-requires-evaluation/
Read more

SecObserve: Simplified Vulnerability and License Management for CI/CD Pipelines

 SecObserve is an open-source tool for managing vulnerabilities and licenses in software development and cloud environments. It integrates various vulnerability scanners into CI/CD pipelines using GitLab CI templates and GitHub Actions for streamlined setup. It offers a centralized dashboard for assessing and reporting vulnerabilities, with tools for filtering, sorting, and evaluating results. SecObserve supports automation and manual assessments to focus on resolving critical issues.  https://github.com/MaibornWolff/SecObserve/tree/dev
Read more

OWASP ZAP 2.16.0 Introduces Key Updates and Enhancements

 In January 2025, OWASP ZAP released version 2.16.0 with several key updates. A redesigned spidering method improves the crawling of modern web applications, particularly those using JavaScript frameworks. The new version also introduces detachable tabs for a more flexible workspace. The minimum Java version required has been updated to Java 17. Additionally, the Client Spider was introduced to enhance crawling by capturing events, DOM changes, and user interactions in modern web apps. The project welcomed new contributors and updated its roadmap for ongoing development. https://www.zaproxy.org/blog/2025-02-04-zap-updates-january-2025/
Read more