Data has benefits, but the potential dangers that come with its availability make regulation equally important. The Defense Advance Research Projects Agency (DARPA) recently explained that they have a research program in place to establish a common ground.
The agency is using its Brandeis research program as the link between accessibility and privacy. The goal is to make certain information available without stripping away some of the agency’s internal safeguards, FedScoop reports. The method being employed is known as “secure multi-party differential privacy,” which wouldn’t require data decryption, but would still prevent it from being rebuilt from random outputs.
FedScoop adds that the research team is using computers to help monitor data access:
The program is also leaning on machine learning, having a computer turn privacy preferences into actionable decisions about who may and may not have your data.
Dr. John Launchbury, a computer scientist for DARPA, said that the agency’s efforts could lead to significant progress in data protection. “Both of these thrusts demand heavy computer science, but if we are successful, we will be able to accelerate information sharing as we can become confident that our data will only be used for its intended purpose and no other,” Launchbury said last week at DARPA’s Wait, What? conference.
Overall, DARPA wants data to be safe without the implementation of draconian laws. “Rather than compromise between these two, [this] research program aims to build a third option, enabling safe and predictable sharing of data while reliably preserving privacy,” Launchbury added.