A new wave of AI systems is starting to transform the economy. These systems include increasingly capable models, multimodal assistants, and autonomous agents that can reason, use tools, make decisions, and execute complex, multi-step tasks in real-world settings. As these capabilities advance, they introduce novel security, privacy, and safety challenges. Addressing these challenges is essential to building trust in such systems.

Perplexity is establishing the Secure Intelligence Institute to study the security, trustworthiness, and practical defense of leading-edge AI systems. Through a portfolio of intramural research and collaboration with the academic community, SII will advance the frontier of AI security research, translate those advances into concrete improvements in Perplexity's systems, and share knowledge and resources that help strengthen the broader AI ecosystem. This work is informed by our experience operating general-purpose AI systems used by millions of users and thousands of enterprises in both controlled and open-world environments.