Dawn Song

Dawn Song is a professor at University of California, Berkeley, in the Department of Electrical Engineering and Computer Science with research interests in deep learning and security. Song performs research in Artificial Intelligence, Operating Systems and Networking, Security and Programming Systems. Song is associated with the following research centers: Real-Time Intelligent Secure Explainable Systems (RISELab), the Center for Human Compatible Artificial Intelligence (CHAI), Foundations of Resilient CybEr-physical Systems (FORCES), Tsinghua-UC Berkely Shenzhen Institute (TBSI)Tsinghua-UC Berkely Shenzhen Institute (TBSI), Berkeley Artificial Intelligence Research Lab (BAIR)Berkeley Artificial Intelligence Research Lab (BAIR), Partners for Advanced Transit and Highways (PATH)Partners for Advanced Transit and Highways (PATH), Berkeley Institute for Data Sceinces (BIDS) and Berkeley Deep Drive (BDD).

Song and her PhD student Noah Johnson commercialized their DroidBlaze platform and founded Ensighta Security, acquired by FireEyeFireEye in 2012.

Faculty Research Award from IBMIBM, Google and other major tech companies

Song is a computer security specialist who uses theoretical methods to probe the interactions between software, hardware and networks which make computer systems vulnerable to attack or interference. Song investigates underlying patterns in computer system behavior which are applicable across whole classes of security. Using a method for semantic analysis of binary code, such as the machine-readable translation of human-readable programmers’ instructions, from disruptive software, Song is able to identify the common path of logic flow that disruptive software must follow. Through her research Song has shown that software patches intended to fix existing security flaws can be used as a template for algorithms that autonomously generate similar computer programscomputer programs that exploit the flaw or circumvent repair. Song helped to develop an algorithm that can protect sensitive information using cryptography.

Dawn Song is founder and CEO of Oasis Labs, a cloud computing company focused on preserving privacy, that uses blockchain technology . The company was founded in 2018, is based in San Francisco and co-founders include Bobby Jaros, Noah JohnsonNoah Johnson and Raymond Cheng. Their privacy platform called Ekiden is a system that addresses gaps in combining blockchains with Trusted Execution Environments (TEEs).

Alfred P. Sloan Research FellowshipSloan Research Fellowship

Guggenheim FellowshipGuggenheim Fellowship

Dawn Song is a professor at University of California, Berkeley, in the Department of Electrical Engineering and Computer Science with research interests in deep learning and security. Song performs research in Artificial Intelligence, Operating Systems and Networking, Security and Programming Systems. Song is associated with the following research centersresearch centers: Real-Time Intelligent Secure Explainable Systems (RISELab), the Center for Human Compatible Artificial Intelligence (CHAI), Foundations of Resilient CybEr-physical Systems (FORCES), Tsinghua-UC Berkely Shenzhen Institute (TBSI), Berkeley Artificial Intelligence Research Lab (BAIR), Partners for Advanced Transit and Highways (PATH), Berkeley Institute for Data Sceinces (BIDS) and Berkeley Deep Drive (BDD).

MIT Technology ReviewMIT Technology Review TR-35 Award

Dawn Song received a B.S. (1996) from Tsinghua UniversityTsinghua University and an M.S. (1999) from Carnegie Mellon University. Dawn Song earned her Ph.D. from UC Berkeley in 2002. She was an Assistant Professor at Carnegie Mellon University from 2002-2007 before joining UC Berkeley as faculty.

Dawn Song is founder and CEO of Oasis Labs, a cloud computing company focused on preserving privacy, that uses blockchain technology . The company was founded in 2018, is based in San FranciscoSan Francisco and co-founders include Bobby Jaros, Noah Johnson and Raymond Cheng. Their privacy platform called Ekiden is a system that addresses gaps in combining blockchains with Trusted Execution Environments (TEEs).

Dawn Song received a B.S. (1996) from Tsinghua University and an M.S. (1999) from Carnegie Mellon UniversityCarnegie Mellon University. Dawn Song earned her Ph.D. from UC Berkeley in 2002. She was an Assistant Professor at Carnegie Mellon University from 2002-2007 before joining UC Berkeley as faculty.

Faculty Research Award from IBM, GoogleGoogle and other major tech companies

Song is one of the leaders of the Keystone Project an academic initiative led by researchers from UC Berkeley and MIT and includes joint development with Oasis LabsOasis Labs and other institutions. The Keystone Project open-sourced its secure enclave framework in December 2018. Keystone utilizes RISC-V based hardware which will be deployed as real RISC-V chips and is designed to be a general research platform for secure hardware innovation.

Song developed a platform called BitBlaze which analyzes malware and automatically generates a filter to protect against it until a security patch is available. The program can also analyze the patches and produce new malware that exploits vulnerabilities, guiding programmers in making security patches. After a security attack, BitBlaze can defend from a variety of future attacks that target the same vulnerability. Technology from Song’s research has been incorporated into the Google Chrome web browser. Song has collaborated with SymantecSymantec, a security software company.

Song developed a platform called BitBlaze which analyzes malware and automatically generates a filter to protect against it until a security patch is available. The program can also analyze the patches and produce new malware that exploits vulnerabilities, guiding programmers in making security patches. After a security attack, BitBlaze can defend from a variety of future attacks that target the same vulnerability. Technology from Song’s research has been incorporated into the Google ChromeGoogle Chrome web browser. Song has collaborated with Symantec, a security software company.