You tailor guardrails to each partner's policies, regulations, and risk appetite.
You red-team every deployment, testing for jailbreaks, prompt injection, and data leakage.
You lead partner security reviews, architecture reviews, and response to safety incidents.
Hopper builds the leading travel technology platform, powering travel for banks, airlines and travel companies globally. It is a well-funded startup with small dynamic teams, an entrepreneurial culture, and over 120 million app downloads.
Perform Disengagement Validation by reviewing AI attacker decisions to ensure each proposed action is appropriate and safe given the specific application context.
Unblock AI authentication hurdles by providing necessary context, solving CAPTCHAs, or utilizing alternative avenues to establish access.
Guide the AI during the reconnaissance phase and ensure Attack Data Classification is accurate and actionable.
At Armadin, we are a group of engineers and researchers building autonomous proactive security for the AI era. Led by Kevin Mandia, our team includes researchers from top tech companies and universities, reinventing security.
Design and lead red teaming campaigns to identify attack vectors in high-risk model capabilities.
Develop jailbreak techniques and document them for reproducibility and knowledge sharing.
Test agent systems for prompt injection vulnerabilities involving untrusted inputs.
We support AI safety testing projects, focusing on identifying and addressing risks before AI systems are deployed. We are a team of experts working on security, fairness, and domain-specific safety.