--- title: How to Break AI Systems (Before Someone Else Does) -... description: A conference talk covering AI red teaming techniques and risks associated with large language models. The speaker explains the fundamental architectural flaw... canonical: https://daily.dev/posts/how-to-break-ai-systems-before-someone-else-does---gary-lopez---ndc-toronto-2026-6rsgoot0b twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: How to Break AI Systems (Before Someone Else Does) - Gary Lopez - NDC Toronto 2026 | daily.dev og:description: A conference talk covering AI red teaming techniques and risks associated with large language models. The speaker explains the fundamental architectural flaw... og:url: https://daily.dev/posts/how-to-break-ai-systems-before-someone-else-does---gary-lopez---ndc-toronto-2026-6rsgoot0b og:image: https://api.daily.dev/og/posts/6rsgOoT0B.png og:image:alt: How to Break AI Systems (Before Someone Else Does) - Gary Lopez - NDC Toronto 2026 og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # How to Break AI Systems (Before Someone Else Does) - Gary Lopez - NDC Toronto 2026 **[NDC Conferences](https://daily.dev/sources/ndc)** · 36 min read · 0 upvotes · 0 comments ## Summary A conference talk covering AI red teaming techniques and risks associated with large language models. The speaker explains the fundamental architectural flaw in LLMs — the inability to distinguish between data and code — which enables jailbreaks and prompt injections. Direct prompt injection techniques are covered including competing objectives (getting models to start with 'sure, here's how to') and mismatch generalization (using Base64, ROT13, or chemical notation to bypass safety filters). The Crescendo attack is explained as a gradual escalation technique that still works against frontier models. Indirect prompt injection via emails, PDFs, and web content is also discussed. Broader risks include model alignment removal via fine-tuning, agentic escalation in multi-agent military simulations, and nation-state exploitation of AI systems. Mitigations covered include the 'spotlighting' family of techniques (delimiting, data marking, and encoding user inputs) plus input/output filters like toxic content detectors, prompt injection classifiers, and fact-checking guards. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [The New Frontier of Cyber Threats: Unpacking Prompt Injection, Model Poisoning and Adversarial Attacks in AI Security](https://daily.dev/posts/the-new-frontier-of-cyber-threats-unpacking-prompt-injection-model-poisoning-and-adversarial-attac-cjvycddit) · Security Boulevard · 0 upvotes · 0 comments - [Indirect Prompt Injection remains a fundamental security challenge for AI](https://daily.dev/posts/indirect-prompt-injection-remains-a-fundamental-security-challenge-for-ai-xnrdac9em) · Brave · 1 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm), [#ai-security](https://daily.dev/tags/ai-security), [#prompt-injection](https://daily.dev/tags/prompt-injection), [#red-teaming](https://daily.dev/tags/red-teaming) [View this post on daily.dev](https://daily.dev/posts/how-to-break-ai-systems-before-someone-else-does---gary-lopez---ndc-toronto-2026-6rsgoot0b) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"How to Break AI Systems (Before Someone Else Does) - Gary Lopez - NDC Toronto 2026","url":"https://daily.dev/posts/how-to-break-ai-systems-before-someone-else-does---gary-lopez---ndc-toronto-2026-6rsgoot0b","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/how-to-break-ai-systems-before-someone-else-does---gary-lopez---ndc-toronto-2026-6rsgoot0b"},"datePublished":"2026-07-28T11:55:29.134Z","dateModified":"2026-07-28T11:55:50.683Z","description":"A conference talk covering AI red teaming techniques and risks associated with large language models. The speaker explains the fundamental architectural flaw...","image":"https://i.ytimg.com/vi/9AG9lUdCgdY/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/9AG9lUdCgdY/sddefault.jpg","isAccessibleForFree":true,"articleSection":"NDC Conferences","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"NDC Conferences","logo":"https://media.daily.dev/image/upload/s--SiIyojEj--/f_auto/v1748760369/logos/ndc","url":"https://daily.dev/sources/ndc"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/how-to-break-ai-systems-before-someone-else-does---gary-lopez---ndc-toronto-2026-6rsgoot0b","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"llm,ai-security,prompt-injection,red-teaming","timeRequired":"PT36M","video":{"@type":"VideoObject","name":"How to Break AI Systems (Before Someone Else Does) - Gary Lopez - NDC Toronto 2026","description":"A conference talk covering AI red teaming techniques and risks associated with large language models. The speaker explains the fundamental architectural flaw...","thumbnailUrl":"https://i.ytimg.com/vi/9AG9lUdCgdY/sddefault.jpg","uploadDate":"2026-07-28T11:55:29.134Z","duration":"PT36M","url":"https://api.daily.dev/r/6rsgOoT0B","embedUrl":"https://www.youtube.com/embed/9AG9lUdCgdY"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"NDC Conferences","item":"https://daily.dev/sources/ndc"},{"@type":"ListItem","position":3,"name":"How to Break AI Systems (Before Someone Else Does) - Gary Lopez - NDC Toronto 2026"}]} ```