December Timeline for EU AI Proposal Remains on Target
The European Commission remains on target to propose artificial intelligence legislation in December, despite COVID-19 complications, EC Team Leader-AI Irina Orssich said Wednesday (see 2002190004). “This might still happen,” she said during an Information Technology and Innovation Foundation livestream.
Sign up for a free preview to unlock the rest of this article
Timely, relevant coverage of court proceedings and agency rulings involving tariffs, classification, valuation, origin and antidumping and countervailing duties. Each day, Trade Law Daily subscribers receive a daily headline email, in-depth PDF edition and access to all relevant documents via our trade law source document library and website.
Europe has the assets to build a strong AI ecosystem but trails the U.S. and China on adoption, commercialization and investment, said ITIF Center for Data Innovation senior policy analyst Eline Chivot. EU members agree a regulatory framework is needed to promote innovation and access to data, she said.
Public consultation is expected to close May 31, said Orssich. An event for consultation is expected in mid-September followed by targeted consultations with experts about provisions, she said: “The timeline hasn’t changed.” Some of the ongoing public consultation can be done remotely, she said. Social distancing has allowed officials to think through the proposal and consult. “At this point, it’s not the worst way of working,” she said, noting things could change in the next few weeks.
The EU asked Google about combining deep-learning data to research COVID-19 and formulate predictions, said Google Manager-Public Policy Sylwia Giepmans-Stepien. There are costs to not using AI, she said, noting AI isn’t a “miracle solution” but has benefits. The EU is in a good position to play a lead role, she said.
COVID-19 highlights why digital technology is essential, said European Tech Alliance Chair Magdalena Piech. The EC released an AI white paper in February setting an approach based on excellence and trust that stresses high-risk AI systems must be transparent and under human control. Defense of European industry is very visible in the EC white paper, she said, hoping it will stay like that throughout the legislative process.
COVID-19 exposed the EC white paper to criticism, said Chivot, citing limitations of relying on EU data sets to develop something like a vaccine. Broader data sets are needed, she said, noting EU members aim to boost “data sovereignty.” The EU is losing the AI race, despite all the data in the region, she said.
Orssich agreed EU members would benefit from access to broader data sets. She noted members are strong in certain fields: industrial data, for instance.
Some member states pretend to be functional or super advanced with AI, when the reality is they’re not, said Lithuania Vice-Minister-Ministry of the Economy and Innovation Elijus Civilis. Countries need to collaborate honestly, he said, while calling for experimentation, not just planning and research.
The purpose of the white paper was to create as much trust as possible and avoid creating too much administrative burden, said Orssich. People understand the importance of digital tools, especially people with less connection to other humans during the health crisis, she said.