Industry Groups Urge AI Data-Sharing; EPIC, Microsoft Suggest Anonymous Data
The administration should promote data-sharing when updating national artificial intelligence strategy (see 1812040056), industry groups and Amazon said in comments to the Networking and Information Technology Research and Development Program posted last week. Microsoft and the Electronic Privacy Information…
Sign up for a free preview to unlock the rest of this article
Export Compliance Daily combines U.S. export control news, foreign border import regulation and policy developments into a single daily information service that reliably informs its trade professional readers about important current issues affecting their operations.
Center (EPIC) voiced support for “de-identification” techniques for anonymous data gathering and sharing. Access to large data sets is essential for Al and machine learning research and development, Amazon said. The e-commerce platform also argued against policies and regulations that might “hamper” tech R&D. The Information Technology Industry Council called data the “gasoline that fuels AI engines,” cautioning that data and privacy concerns must be considered. Sharing data would allow industry to better train algorithms, ITI said. In 2018, the U.S. didn’t properly fund AI R&D, the Information Technology and Innovation Foundation’s Center for Data Innovation said, citing better-financed competition from China, France and the U.K. Exposing source code for AI technology wouldn’t be useful, ITIF said, arguing “transparency guarantees neither accurate nor unbiased results.” The Software & Information Industry Association highlighted passage of the National Quantum Initiative Act, which President Donald Trump signed last week, authorizing $1.2 billion over five years for quantum activities across the federal government. People have a right to transparency, including data on human decision-making and the identities of groups behind the technology, EPIC said, offering 12 core principles endorsed by more than 200 experts and 50 NGOs. No group should be able to maintain “secret” profiling systems, and groups should be obligated to terminate a system if “human control of the system is no longer possible,” EPIC said. Echoing comments from EPIC, Microsoft backed de-identification data-sharing, or methods that preserve confidentiality, privacy and security. “However, AI systems that are used in contexts that involve people would need access to data about people to make informed decision[s],” Microsoft said.