0Din.ai

Public Disclosure: Amazon Rufus Guardrail Jailbreak via ASCII Integer Encoding https://lnkd.in/gVqdPKkU (0din.ai) The 0Din team discovered this vulnerability. More information is available in the tandem blog at: https://lnkd.in/gxUV-KUT (0din.ai blog)

Public Disclosure: Amazon Rufus Guardrail Jailbreak via ASCII Integer Encoding https://lnkd.in/gVqdPKkU (0din.ai) The 0Din team discovered this vulnerability. More information is available in the tandem blog at: https://lnkd.in/gxUV-KUT (0din.ai blog)

??Breaking: Data poisoning isn’t just theory anymore! ?? @elder_plinius github repository was fine-tuned into Deepseek DeepThink (R1) and has “liberated” an AI from its own guardrails, no internet connection required. ?? Dive into my new blog, “Poison in the Pipeline: Liberating Models with Basilisk Venom,” to the breakdown of this technique and why it changes everything for AI. ?? Kudos to Dominick Romano for highlighting the risk of data poisoning back in July. Your foresight enabled me to remain vigilant and identify it promptly when it occurred. https://lnkd.in/eb5BCiZG #AI #DataPoisoning #Cybersecurity

Check out this weeks 0Din #GenAI #BugBounty Submissions! Also make sure you read the #0Din blog that was published on Monday for #ChatGPT-4o Guardrail #Jailbreak by Marco Figueroa

?? 0Din Blog Alert: New insights on AI jailbreaks! ?? Just released: a deep dive into how 0Din researchers discovered how to leverage hex encoding to bypass ChatGPT-4o’s security measures and generate exploit code for CVEs. This simple technique opens up a whole new perspective and conversation on AI risks and cybersecurity. In this blog post, we cover: 1?? How simple techniques are outsmarting AI guardrails. 2?? The sneaky way hex encoding fools some language models into producing exploit code. 3?? A step-by-step walkthrough of why this technique works. Check out the complete analysis and explore what this means for the future of AI and security: https://lnkd.in/eQfJiDid

??New Vulnerabilities Submitted by Researchers to the 0Din GenAI Portal?? ??? This isn’t just about the numbers, it’s about precision. Every vulnerability submitted is vetted. When submissions don't meet the CBRN (Chemical, Biological, Radiological, and Nuclear) criteria or a specific technique listed on our Boundary List, we collaborate with researchers to refine and strengthen the results, ensuring the #0Din community stays empowered. ?? Why does this matter?? AI is advancing at an accelerated pace. It's crucial to distinguish between what's genuinely possible and what's merely speculation, especially for those who believe AI poses little to no risk. Understanding the real-world implications is key to recognizing the potential dangers AI can bring. ?? If you're working in the GenAI space, this video gives an inside look at the vulnerabilities we're receiving. ?? #AI #Cybersecurity #LLM #GenAI

??0Din GenAI Bug Bounty Update?? ?? At 0Din, we're not just building technology we're crafting the future of GenAI security. We believe that our dedicated researchers are the cornerstone of safeguarding tomorrow's AI systems. ?? Their relentless pursuit of vulnerabilities are critical to ensuring that the AI of the future is not only powerful but also secure, resilient, and trusted by all. ?? Together, we’re shaping an ecosystem where innovation thrives safely. In just ONE WEEK, we’ve hit 32 vulnerability submissions! ?? This is just the beginning, and we’re excited to see what the next weeks will bring! ?? #AI #BugBounty #cybersecurity #LLM #GenAI #0Din

?? ODIN Portal Launch Update ?? ?? In just the first 48 hours since launch, we’ve received 10 GenAI vulnerabilities submitted by our incredible community! ??This is just the beginning! The excitement and engagement are already exceeding expectations, and we can’t wait to see how much more we can accomplish together. ??? Stay tuned for more updates as we continue pushing the boundaries of GenAI security! ?? Have insights or discoveries? Keep them coming, and let’s make GenAI safer for everyone. #GenAI #AIsecurity #LLM #BugBounty #Cybersecurity

?? 0Din Blog: Revolutionizing Bug Bounty Hunting for GenAI Security ?? I’m thrilled to introduce the 0din Portal, a revolutionary platform designed specifically for GenAI Security. This isn’t just an upgrade—it’s a game changer for bug bounty hunters everywhere. Here’s why: 1?? AI-Powered Insights: Say goodbye to hours of manual bug tracking. Our portal leverages cutting-edge GenAI to enhance triaging submissions, making the process faster and more effective than ever. 2?? Streamlined Collaboration: We’re bridging the gap between security researchers and organizations by offering real-time collaboration tools. This means quicker fixes and more robust security for everyone. 3?? Tailored to GenAI Security: Traditional platforms weren’t built for the complexities of GenAI—we were. We understand that GenAI security is the future of bug bounties. ??♂? We’re making security smarter, faster, and more collaborative. If you’re passionate about the future of bug bounties or cybersecurity, this is the portal you’ve been waiting for. ?? Ready to explore the future? Check out more details here: https://lnkd.in/eMiM4UHU P.S. Share this with someone who needs to know about the future of bug bounties. ?? #BugBounty #Cybersecurity #GenAI #AI #LLM #Infosecurity #0DinAI

??Recent LLM Bug Hunting Insights?? ?? With the release of several new large language models (LLMs), our researchers at 0Din have noticed a significant increase in IP blocking during bug bounty hunting. ?? 0Din has you covered! ?? To ensure our community researchers can continue their work even when IPs are temporarily blocked, anyone who submits a vulnerability to 0Din will receive a free 3-month license to Mozilla’s VPN service. ?? 500+ servers in 30+ countries ?? Connect up to 5 devices ?? No bandwidth limits or throttling ?? At 0Din, we’re committed to providing value to our community and enhancing the tools you need for success. Hope this helps, repost??? for others! ?