Search results

The European Commission recently published the policy recommendations of its "High-Level Expert Group on Artificial Intelligence": a heavily anticipated document, particularly in the context of the stated ambition of the new Commission President to regulate in that area. This article argues that these recommendations have significant deficits in a range of areas. It analyses a selection of the Group's proposals in context of the governance of artificial intelligence more broadly, focusing on issues of framing, representation and expertise, and on the lack of acknowledgement of key issues of power and infrastructure underpinning modern information economies and practices of optimisation.

Cybersecurity covers the broad range of technical and social issues that must be considered to protect networked information systems. The importance of the concept has increased as so many government, business, and day-to-day activities globally have moved online. It has been increasingly referred to in both academic and mainstream publications since 2003, in fields including software engineering, international relations, crisis management and public safety, slowly overtaking more technical terms such as computer/system/data security (popular in the 1970s/1980s) and information security (popular from the mid 1990s). But its strong association with national security and defence agencies, and disconnection from social science notions such as place, have led to concerns of inappropriate cyber securitisation of government programmes.

AbstractConcerns around machine learning's societal impacts have led to proposals to certify some systems. While prominent governance efforts to date center around networking standards bodies such as the Institute of Electrical and Electronics Engineers (IEEE), we argue that machine learning certification should build on structures from the sustainability domain. Policy challenges of machine learning and sustainability share significant structural similarities, including difficult to observe credence properties, such as data collection characteristics or carbon emissions from model training, and value chain concerns, including core‐periphery inequalities, networks of labor, and fragmented and modular value creation. While networking‐style standards typically draw their adoption and enforcement from functional needs to conform to enable network participation, machine learning, despite its digital nature, does not benefit from this dynamic. We therefore apply research on certification systems in sustainability, particularly of commodities, to generate lessons across both areas, informing emerging proposals such as the EU's AI Act.