1.0Cyber Security Newshttps://cybersecuritynews.comNew Echo Chamber Attack Jailbreaks Most AI Models by Weaponizing Indirect Referencesrich600338<blockquote class="wp-embedded-content" data-secret="LPQKnT1CRb"><a href="https://cybersecuritynews.com/echo-chamber-attack/">New Echo Chamber Attack Jailbreaks Most AI Models by Weaponizing Indirect References</a></blockquote><iframe sandbox="allow-scripts" security="restricted" src="https://cybersecuritynews.com/echo-chamber-attack/embed/#?secret=LPQKnT1CRb" width="600" height="338" title="“New Echo Chamber Attack Jailbreaks Most AI Models by Weaponizing Indirect References” — Cyber Security News" data-secret="LPQKnT1CRb" frameborder="0" marginwidth="0" marginheight="0" scrolling="no" class="wp-embedded-content"></iframe><script type="text/javascript">
/* <![CDATA[ */
/*! This file is auto-generated */
!function(d,l){"use strict";l.querySelector&&d.addEventListener&&"undefined"!=typeof URL&&(d.wp=d.wp||{},d.wp.receiveEmbedMessage||(d.wp.receiveEmbedMessage=function(e){var t=e.data;if((t||t.secret||t.message||t.value)&&!/[^a-zA-Z0-9]/.test(t.secret)){for(var s,r,n,a=l.querySelectorAll('iframe[data-secret="'+t.secret+'"]'),o=l.querySelectorAll('blockquote[data-secret="'+t.secret+'"]'),c=new RegExp("^https?:$","i"),i=0;i<o.length;i++)o[i].style.display="none";for(i=0;i<a.length;i++)s=a[i],e.source===s.contentWindow&&(s.removeAttribute("style"),"height"===t.message?(1e3<(r=parseInt(t.value,10))?r=1e3:~~r<200&&(r=200),s.height=r):"link"===t.message&&(r=new URL(s.getAttribute("src")),n=new URL(t.value),c.test(n.protocol))&&n.host===r.host&&l.activeElement===s&&(d.top.location.href=t.value))}},d.addEventListener("message",d.wp.receiveEmbedMessage,!1),l.addEventListener("DOMContentLoaded",function(){for(var e,t,s=l.querySelectorAll("iframe.wp-embedded-content"),r=0;r<s.length;r++)(t=(e=s[r]).getAttribute("data-secret"))||(t=Math.random().toString(36).substring(2,12),e.src+="#?secret="+t,e.setAttribute("data-secret",t)),e.contentWindow.postMessage({message:"ready",secret:t},"*")},!1)))}(window,document);
/* ]]> */
</script>
https://i1.wp.com/blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEiESdk4InVtnWkxkP79uhKHo9z6gv9bumZh1IVeHcl04oryiJMvdV59joT2VSzaZ3StEo-jGX9KAVwrCJFZ04vqjur16SKShqfHXp6nrjrYwHeBL_50BORH4o6nqY5-V1CjZRDvtTgEfvFp9IgfK75el_FduXI6uChyphenhyphenUlJEmKshSzf8Leg624qXWSGM8MQr/s16000/Echo%20Chamber%20Attack.webp?w=1600&resize=1600,900&ssl=11600900"Echo Chamber Attack," this method leverages context poisoning and multi-turn reasoning to guide models into generating harmful content without ever issuing an explicitly dangerous prompt.