https://arshimtiaz.com/tags/stride/ Recent content in Stride on Arsh Imtiaz Hugo en Sun, 06 Sep 2026 19:57:52 +0100 https://arshimtiaz.com/posts/stride-ai-threat-modelling/ Sun, 19 Jul 2026 00:00:00 +0100 https://arshimtiaz.com/posts/stride-ai-threat-modelling/ <figure><a class="lightgallery" href="https://arshimtiaz.com/images/stride-ai-threat-modelling-hero.svg" title="Hero diagram showing a poisoned document flowing through an AI assistant into multiple STRIDE impacts" data-thumbnail="/images/stride-ai-threat-modelling-hero.svg" data-sub-html="<h2>The model may work as designed while the trust chain around it is compromised.</h2>"> <img class="lazyload" src="https://arshimtiaz.com/svg/loading.min.svg" data-src="https://arshimtiaz.com/images/stride-ai-threat-modelling-hero.svg" data-srcset="https://arshimtiaz.com/images/stride-ai-threat-modelling-hero.svg, https://arshimtiaz.com/images/stride-ai-threat-modelling-hero.svg 1.5x, https://arshimtiaz.com/images/stride-ai-threat-modelling-hero.svg 2x" data-sizes="auto" alt="Hero diagram showing a poisoned document flowing through an AI assistant into multiple STRIDE impacts" /> </a><figcaption class="image-caption">The model may work as designed while the trust chain around it is compromised.</figcaption> </figure> <!-- This post deliberately treats the model as one component in a wider system. --> <p>An AI assistant can follow every instruction it was given and still help an attacker compromise the organisation.</p>