AI Safety Comparison: OpenAI o3-mini vs. Deepseek-R1 VirtueAI conducted an in-depth red-teaming evaluation of two leading AI models to assess their safety, bias, privacy protections, and robustness. Key findings: 1. o3-mini demonstrates stronger privacy safeguards and fairness

New paper & model release! Excited to introduce DuoGuard: A Two-Player RL-Driven Framework for Multilingual LLM Guardrails, showcasing our new DuoGuard-0.5B model. - Model: huggingface.co/DuoGuard/DuoGu… - Paper: arxiv.org/abs/2502.05163 - GitHub: github.com/yihedeng9/DuoG… Grounded in a

We’re partnering with Glean to adapt Virtue AI’s pioneering research in AI security, including content moderation, guardrails, and red teaming to Glean’s enterprise customers. Find out more: bit.ly/3CZs6I6

Toward Trustworthy Generative Foundation Models (GenFMs) 🚀 🎇After six months of hard work and thanks to the efforts of the entire team, our report on the trustworthiness of generative foundation models (GenFMs) has finally been released. 💡In this work, we: -Developed a

Can Reasoning Improve Safety & Security? Red-Teaming Analysis for Claude 3.7 🚀 Claude 3.7 Sonnet Thinking: A New Era of Hybrid Reasoning? Anthropic's latest release introduces a Thinking mode, letting users switch between rapid responses and step-by-step reasoning. But does

Virtue AI just released our red-teaming analysis of OpenAI’s GPT-4.5 in comparison with Paul Jankura’s Claude 3.7! We tested them on safety, security, hallucination, regulatory compliance, codeGen vulnerabilities, and more. Here’s what we found... (1/9)

Virtue AI is honored to be recognized by Intel's new CEO, Lip-Bu Tan, during his opening keynote at #IntelVision. Wishing him great success as he leads Intel into its next exciting chapter. AI security & safety have become the critical last mile for AI applications. At Virtue

Exciting!! Huge congratulations to Chejian Xu Wei Ping and the great NVIDIA team!! Looking forward to the next exciting milestone!!

Join Virtue AI Co-founder Sanmi Koyejo for a live webinar on why protecting your AI apps isn’t just about safety—it’s the key to faster deployment and growth. 📅 April 24 | 🕙 10 AM PT | 💻 Virtual In this session, we’ll cover: ✅ Why traditional security tooling falls short

We’ve raised $30M in Seed + Series A funding led by Lightspeed and Walden Catalyst Ventures, with participation from Prosperity7 Ventures, Factory, Osage University Partners (OUP), Lip-Bu Tan, Chris Re, and more. Virtue AI is the first unified platform for securing AI across

We are super excited to empower developers to focus on their goal of building innovative AI applications; we’ll take care of safety and security! What an awesome ride with Bo Li Bo Li, Sanmi Koyejo, Dawn Song and the whole Virtue AI team!

📷Come and check our #ICLR2025 poster 𝗦𝗮𝗳𝗲𝗪𝗮𝘁𝗰𝗵!!🔥 Today April 25th 3:00 pm -5:30 pm 📷 at Poster Session Hall 3 #547📷📷

🚨 3 days out from our live webinar on the EU AI Act hosted by Sanmi Koyejo and Jan Eißfeldt! Register now: us06web.zoom.us/webinar/regist… ⬇️ Details below

VirtueAgent provides the first systematic guardrails for general AI agents!! Super exciting work such that we can rest assured and let our agents handle things for us!👍

🚨 Introducing VirtueGuard Code: Real-time vulnerability detection for AI-generated code. As coding assistants like Cursor and GitHub Copilot become standard in development workflows, it’s critical to ensure that generated code meets security standards. VirtueGuard Code is

Congrats to our partners at Glean on their first-ever hashtag #GleanGO conference! 🎉 We’re honored to be part of their security and governance ecosystem, helping power trusted AI across the enterprise.

Very timely Guard Agent to ensure access control for general agents in different domains!

Autonomous AI agents are rapidly being deployed across industries, from web browsing copilots to code-writing assistants and enterprise workflow agents. But these systems come with a new class of security risks that traditional guardrails and red teaming aren’t equipped to

Safety & security definitions are domain-specific in most cases -- We provide the first domain-specific, and policy-grounded guardrail benchmark! Exciting to enter the stage of nuanced guardrail protection for foundation models and AI applications!

🛡️ VirtueGuard is LIVE on Together AI 🚀 AI security and safety model that screens input and output for harmful content: ⚡ Under 10ms 𝗿𝗲𝘀𝗽𝗼𝗻𝘀𝗲 🎯 𝟴𝟵% 𝗮𝗰𝗰𝘂𝗿𝗮𝗰𝘆 vs 76% (AWS Bedrock) 🧠 𝗖𝗼𝗻𝘁𝗲𝘅𝘁-𝗮𝘄𝗮𝗿𝗲 - adapts to your policies, not just keywords 👇