When multiple parties attempt to collaborate, they often end up having to compromise the confidentiality of their private information. Cryptographic primitives like secure multiparty computation and obfuscation are aimed at mitigating this situation. We view many such problems as part of a unifying paradigm that we call Active Information: various parties in a system can have access to prescribed actions of the information in the system, but nothing more.The first focus area of the project is secure multiparty computation, which lets mutually distrusting parties emulate a commonly trusted party, by means of a protocol. We identify and investigate crucial properties required for this to be useful in practice: "network-awareness" addresses issues that arise in a realistically complex environment, and a new class of "context-awareness" features are vital in integrating secure multiparty computation into a larger application. We also derive insight into the fundamental nature of various multiparty computation tasks by classifying them into complexity classes. Another focus of this project is on "secure active objects" that offer far more flexibility in their use, compared to protocols. We study well-known (but little understood) concepts like obfuscation, as well as new tools we develop like "remotely primed obfuscations," and "active encryptions," and also features like anonymity and unlinkability that are important for applications. Secure active objects also serve as a platform to integrate the study of hardware-based security solutions into the theory of cryptography. The tools developed span a wide spectrum, ranging from low-level algebraic techniques used in constructions, to high-level frameworks for conceptualizing security definitions. Towards bringing the subtle concepts from theoretical cryptography to a broader audience of current and future practitioners, two online resources are being developed: (1) a wiki-based tutorial/reference website, and (2) a blog aimed at students and a non-specialist audience.